• Global smart glasses shipments grew 44.4% in 2025, driven by lightweight AI-first wearables rather than bulky headsets.
• The market is divided into three categories: AR glasses with displays, screen-free AI eyeglasses for productivity, and audio smart glasses for calls and music.
• 2026 marks a turning point due to major tech company involvement, improved hardware like Bluetooth 5.3 and 4-microphone arrays, and enhanced features like real-time translation and AI meeting transcription.
• Key considerations for purchasing include battery life, prescription lens compatibility, camera-free designs for enterprise use, and lightweight materials like titanium.

• Google is preparing to launch three new smart glasses products this year, including an advanced AR setup called Project Aura and wireless AI glasses developed with Samsung, Gentle Monster, and Warby Parker.
• These upcoming glasses aim to compete with Meta's smart glasses, featuring AI capabilities powered by Gemini that integrate with existing phone and watch apps.
• Prototypes demonstrated AI-driven features like real-time translation, on-screen captions, and AI-generated widgets, with a focus on lightweight designs and potential prescription support.
• While official names and prices are not yet released, these glasses are expected to be available in the fall, with a dual-display model potentially arriving later.
• Google and Samsung acknowledge public apprehension regarding privacy for camera-enabled glasses and plan to provide more details on data privacy at a fall event.

• Samsung Electronics is expected to launch its new smart glasses series in autumn, potentially challenging Meta's market dominance.
• These display-free smart glasses, developed with Gentle Monster and Warby Parker, will integrate with Google's Gemini AI for features like navigation and translation.
• Market observers anticipate pricing will be a crucial factor in this competitive sector, with Meta and Chinese companies already offering lower-cost options.
• The global intelligent eyewear market saw significant growth, particularly in display-free and AR glasses, suggesting consumer preference for lighter, more conventional-feeling eyewear.

• Smartglasses are shifting towards lighter designs, according to a July 2026 analysis by Display Dynamics.
• This trend occurs as the microdisplay industry faces new challenges.
• The article discusses Omdia's recommendations and analyst services, accessible to subscribers.

• In 2026, the iPhone 7 is no longer supported by Apple, having been capped at iOS 15 since 2022.
• This lack of software updates leads to security vulnerabilities, app compatibility issues, and diminished performance due to aging hardware.
• Battery life is poor, storage is limited, and camera quality is significantly lower than modern devices.
• Alternatives like the iPhone SE (3rd Gen) and iPhone 12 offer better performance, 5G connectivity, and continued software support, making them more advisable purchases.

• The iPhone 7, released in September 2016, is technically functional for basic tasks in 2026, but is not recommended as a primary smartphone.
• Apple ended official software updates for the iPhone 7 with iOS 15.8 in October 2023, meaning no further security patches or bug fixes are issued.
• This lack of updates leads to app incompatibility, security vulnerabilities, and issues with modern authentication standards and online services.
• The iPhone 7 might be considered for use as a secondary device, in controlled educational or kiosk environments, or for legacy iOS development testing.

• As of early 2026, the iPhone 7 is minimally functional for basic tasks like calls, texts, and light web browsing but is no longer secure, supported, or future-proof.
• It stopped receiving iOS updates with iOS 15.8.7 and lacks modern security patches, preventing it from running newer apps.
• While affordable, its hardware is significantly aged, with many refurbished units having degraded batteries or cosmetic issues.
• The iPhone 7 is only recommended for dedicated offline use, very low-demand senior users, or as a hardware learning tool, due to security and carrier compatibility limitations.

• Electric vehicle (EV) charging technology is set to evolve significantly over the next decade, driven by increasing EV adoption and government initiatives.
• Key innovations include ultra-fast public charging, smart charging with Vehicle-to-Grid (V2G) capabilities for grid stability, and wireless charging.
• The goal is to create a widespread, interoperable charging network integrated with renewable energy sources, making EV ownership more convenient and sustainable.
• These advancements aim to reduce charging times, offer bidirectional energy flow, and expand charging infrastructure to be as common as current petrol stations.

• In 2026, electric vehicle (EV) charging infrastructure in Europe is largely meeting demand, with EV sales growing 26.2% year-over-year in Q1.
• BYD's Flash Charging technology has launched in Europe, offering up to 1 MW per cable with a 10-70% charge in approximately 5 minutes, and has committed to deploying 3,000 stations across Europe within a year.
• Megawatt Charging System (MCS) corridors for trucks are taking shape in Europe, with the first public MCS charger operational in Sweden, and standardization efforts are progressing.
• Battery-buffered charging solutions are becoming common for high-power deployments, offering economic and practical advantages for grid limitations.
• Charging stops are evolving into destination points, with amenities like lounges, restaurants, and play areas becoming key differentiators for charging networks.

• Charge point operators (CPOs) face declining driver satisfaction despite infrastructure growth, with "charge anxiety" now surpassing "range anxiety" due to cost and payment issues.
• To scale EV charging networks by 2026, CPOs must prioritize creating a frictionless driver experience by addressing payment, pricing, and session flow issues.
• Smarter energy management on-site is critical to handle increased demand from DC fast-charging, manage grid constraints, and reduce operational costs.
• Utilizing collected operational data is essential for informing business growth, optimizing site selection, and improving revenue and efficiency.

• In Q2 2026, the US added 4,382 new DC fast-charging ports across 806 stations, a slight decrease from the previous year.
• The total number of public DC fast-charging ports reached 77,776 at 14,514 locations.
• Despite an increase in charging sessions to 46 million, the utilization rate remained stable at 15.8% due to expanded charging capacity.
• Tesla led new port deployments with 1,185, while Walmart, ChargePoint, and Red E also saw significant installations.
• The average price for DC fast charging was $0.538 per kWh, with high-power chargers becoming the new default for non-Tesla networks.

• In Q2 2026, the US public DC fast charging network saw a 10% year-over-year decrease in new port installations, with 4,382 added.
• This slowdown indicates a shift from rapid expansion to a focus on reliability, customer experience, and profitability, marking the industry's entry into "Charging 2.0."
• Ultra-fast charging (250 kW+) is becoming the norm, with 72% of new ports offering this speed, and charging sessions increased by 29% year-over-year.
• Tesla led new installations with 27% of all new ports, while California, Texas, Florida, Illinois, and New York accounted for about 40% of new stations.
• The accelerating adoption of the NACS connector standard is becoming crucial for charge point operators to remain competitive.

• Space Exploration Technologies Corp. (SPCX) has a consensus price target of $236.16 from 29 analysts.
• The highest price target is $800 from Raymond James, while the lowest is $115 from CFRA.
• Most recent ratings on July 16, 2026, from Piper Sandler, Needham, and Evercore ISI Group average $212, suggesting a potential 70.77% upside.
• Piper Sandler's latest rating on July 16, 2026, initiated a neutral stance with a price target of $156.00, implying a 25.66% upside within 12 months.

• ViaSat (VSAT) reported its Q4 2026 earnings on May 28, 2026, with an actual earnings per share (EPS) of -$0.02.
• This figure missed analysts' expectations of $0.28 by 107.14%.
• The company reported the same EPS of -$0.02 in the same quarter last year.
• ViaSat's next earnings call is scheduled for August 4, 2026.

• 37 Wall Street analysts have issued ratings for SpaceX stock in the past year, resulting in a consensus rating of "Moderate Buy."
• The average 12-month price target for SpaceX (SPCX) is $234.78, with a high target of $800.00 and a low of $115.00.
• This consensus price target represents an anticipated upside of 89.35% from the current stock price of $123.99.
• The analysts' ratings include 1 sell, 8 hold, 24 buy, and 4 strong buy recommendations.

• Amazon Leo, a LEO satellite internet service from Amazon, is scheduled to launch in Q1 2026 with projected sub-$400 terminals and AWS integration.
• Existing GEO alternatives like Viasat and HughesNet offer lower upfront costs and reliable rural connectivity, though real-world speeds may differ from advertised rates.
• Other competitors include OneWeb and Telesat Lightspeed, focusing on enterprise and business-critical applications with features like symmetrical bandwidth and guaranteed performance.
• The article also highlights that 5G home internet can be a faster and cheaper alternative where available, and notes potential hidden costs associated with satellite internet, such as power consumption and demand surcharges.

• The AI data center industry is experiencing significant growth and profitability, driven by the increasing demand for computing power to support artificial intelligence.
• Key players in this sector include data center REITs like Equinix and Digital Realty Trust, infrastructure providers such as Vertiv and Broadcom, and AI chip makers like Nvidia and AMD.
• While the market shows strong potential, investors should be cautious due to high stock valuations and the capital-intensive nature of the industry, which is susceptible to interest rate changes and technological shifts.

• New York has become the first U.S. state to implement a statewide moratorium on certain AI data center development projects.
• This action, driven by concerns over energy consumption, land use, and community impact, could signal a broader regulatory trend as other states consider similar legislation.
• While the immediate impact on national data center growth may be limited, the moratorium and potential repeal of tax exemptions are increasing regulatory risk and capital expenditures for developers in New York.
• The development underscores the growing importance of state politics and community dynamics in data center site selection and future infrastructure planning.

• Snapchat Spectacles, video-recording sunglasses, have been released and are available through pop-up vending machines in Venice, California.
• The Spectacles offer a simple, fun experience for capturing 10-second videos with a single button press.
• Videos are transferred to a smartphone via Bluetooth, though syncing can be slow, and then uploaded to Snapchat.
• The glasses feature a circular video format that allows for interactive viewing within the Snapchat app.
• Priced at $129.95, Spectacles aim to be an accessible tool for spontaneous content creation and sharing.

• GoPro's Max 2 360-degree camera bundle is available for $369 at Amazon, $370 at Best Buy and GoPro.
• The bundle includes over $100 worth of accessories, such as an extension pole, extra batteries, a 64GB microSD card, and various mounts.
• The Max 2 camera can record 360-degree 8K video at 30fps or 180-degree 4K video at 60fps.
• It features a six-mic array for audio recording, a 1.82-inch touchscreen, and is waterproof up to 16 feet.

• Filmmaker Werner Herzog discusses his new documentary "Lo and Behold," which examines humanity's adaptation to online life.
• He contrasts his intensive, in-person Rogue Film School with his more accessible MasterClass online course, emphasizing finding one's own voice in filmmaking.
• Herzog expresses skepticism about the necessity of traditional film schools, though he acknowledges their persistence.
• He notes the positive, albeit surprising, reception to "Lo and Behold" from younger audiences accustomed to constant connectivity.

• A collection of the staff's favorite photographs from 2016, culled from various assignments, are presented chronologically.
• The images showcase a diverse range of subjects including technology, events, people, and vehicles.
• Photographers detail their creative process and challenges in capturing these moments, often highlighting unique settings and props.
• The compilation includes photos from events like CES, SXSW, and the Goodwood Festival of Speed, as well as features on products and public figures.

• Walt Mossberg and Nilay Patel reviewed the iPhone 7 and 7 Plus on Ctrl-Walt-Delete.
• The review compared their impressions of Apple's new devices.
• The discussion also touched upon the timing of the iPhone 7 release in relation to the issues with the Samsung Galaxy Note 7.

• A 600-mile road trip to Montreal with an Audi e-tron demonstrated significant improvements in EV charging infrastructure, with only one minor issue encountered.
• The author's recent trip used A Better Route Planner app, which accurately directed them to reliable charging stations, including a Rivian charger with multiple working high-speed chargers.
• Compared to a frustrating trip three years prior, the current charging experience was described as near-flawless, with sessions lasting around 20 minutes and often combined with breaks.
• Data indicates a substantial increase in DC fast chargers nationwide since 2023, with reliability improving from 85% to the mid-90s on Paren's index.

• A blog post reviews DeepWiki, a site that generates documentation for GitHub repositories using AI.
• The author found the generated documentation for mod_blog to be mostly accurate, noting a few minor errors and an issue with its description of system layers.
• Testing with a09, a 6809 assembler, revealed more significant inaccuracies and errors in the generated documentation, particularly concerning error classification, code detection, and data formatting.
• The author concludes that while the tool is less problematic than AI code generation, its current inaccuracies make it unsuitable for unfamiliar or complex codebases.

• The space economy recorded $31.6 billion in private investment across 129 companies in the first half of 2026, exceeding all of 2025 and marking a record-breaking year.
• This surge in investment is largely attributed to SpaceX's recent initial public offering (IPO), which provides a public valuation benchmark for the company.
• SpaceX's IPO filing and subsequent actions, including merging with xAI and acquiring Cursor, indicate a strategic focus on integrating AI and compute capabilities alongside its launch and satellite infrastructure.
• This integrated approach, spanning hardware and software across space, connectivity, and AI, challenges traditional classifications of space companies as solely aerospace and defense entities.

• Boeing CEO Kelly Ortberg will attend the Farnborough Air Show facing competition from Airbus and navigating the political influence of US President Donald Trump.
• Ortberg's agenda includes securing orders, meeting customers, and addressing production speed amid supply chain issues.
• Trump has publicly claimed significant influence over Boeing's sales, suggesting he was instrumental in its recent rebound.
• The political backdrop may influence the timing of aircraft deal announcements, with some airlines potentially waiting for White House engagement to claim shared credit.

• The rapid growth of artificial intelligence is driving a significant expansion of data centers globally, with an estimated 11,800 currently operating.
• These facilities, essential for processing the vast amounts of data required by AI applications, consume substantial amounts of electricity and water, raising concerns about resource sustainability.
• The United States leads in data center numbers, with Virginia, Texas, and California having the most, though geographic distribution is uneven worldwide.
• Expansion faces challenges including access to critical minerals, workforce shortages, energy and water demands, and policy impacts like tariffs on construction materials.

• AI data centers are specialized facilities designed to train and deploy advanced machine learning and large language models (LLMs).
• They differ from traditional data centers by requiring massive power capacity, advanced liquid cooling systems, and dense networking optimized for GPU-intensive computing.
• These centers can consume significantly more power and water than conventional facilities, often necessitating placement near renewable energy sources or in cooler climates.
• Developing AI data centers involves a multidisciplinary approach to site selection, scalable design, thermal and power optimization, and data-driven construction management.

• AI data centers face escalating demands due to advanced AI models like LLMs, necessitating new approaches to power, cooling, and utility integration.
• Extreme power density from AI hardware like GPUs requires upgraded electrical systems and dynamic power management via Integrated Data Center Management (IDCM) platforms.
• Intense thermal loads are driving a shift to advanced liquid cooling solutions, requiring new facility designs and IDCM integration with environmental controls.
• AI data centers strain regional power grids, demanding active partnerships with utilities for planning and grid stability, with data centers potentially offering ancillary services.
• The definition of integrated management is expanding to include external systems like utility grids and renewable energy sources for comprehensive oversight.

• SpaceX scrubbed Starship Flight 13 on Thursday evening due to four Raptor engines failing to ignite at T-zero.
• Elon Musk confirmed the issue, stating two Raptor engines will be replaced, with a probable launch targeting early next week.
• This flight was intended to deploy 20 Starlink V3 satellites and conduct an in-space engine relight, a critical step for future missions.
• The abort occurred during SpaceX's first launch attempt since its stock market debut, increasing public visibility.

• Elorian AI, a new AI research and product lab, has raised $55 million in seed funding.
• The company focuses on advancing multimodal reasoning and visual intelligence, a field led by former Google DeepMind leaders like co-founder Andrew Dai.
• The funding will support research to integrate visual reasoning with language for applications in engineering, robotics, and agriculture.
• Investors include Striker Venture Partners, Menlo Ventures, and Altimeter.

• Between Q1 2024 and Q3 2026, the AI lab market saw 81 funding deals totaling over $240 billion.
• OpenAI, Anthropic, and xAI received the largest funding rounds, with AI lab market funding increasing sharply in 2026.
• Investment focused on frontier models, AI infrastructure, evaluation tools, and AI safety, with companies like Lambda and CoreWeave raising capital for computing capacity.
• AI evaluation emerged as a significant funding category, attracting companies such as Arize AI and Fiddler AI.

• Funding for pure-play AI labs dramatically increased from $30.2 billion in 2024 to $89.1 billion in 2025, and reached $255.9 billion year-to-date in 2026.
• This capital growth is driven by a small number of large, strategic investments in frontier AI labs, with the top three rounds in early 2026 accounting for 84.8% of all funding.
• AI research labs dominate funding, capturing over 70% of capital, and North America is becoming increasingly dominant in the market.
• The market is bifurcating into large-scale model and compute companies and smaller software and tooling companies, with funding heavily concentrated in later-stage rounds.

• Anthropic's Claude 3.5 Sonnet, launched late 2024, is a safety-focused AI that has achieved commercial success.
• The company targets enterprise clients in legal, healthcare, and financial services, offering features like large context windows (500,000 tokens) and regulatory compliance.
• Anthropic is expanding through partnerships with AWS and Google Cloud, opening European offices, and developing agentic workflows for RPA.
• The company has secured over $10 billion in funding and reported a revenue run-rate exceeding $1 billion by the end of 2024.

• AI company Anthropic has secured $65 billion in Series H funding, reaching a post-money valuation of $965 billion.
• The funding, led by Altimeter Capital, Dragoneer, Greenoaks, and Sequoia Capital, will support safety research, expand compute capacity for its AI model Claude, and scale products and partnerships.
• Anthropic reports significant growth in enterprise adoption of Claude and has expanded its compute capacity through agreements with Amazon, Google, and SpaceX.

• GPT-5.6 achieved the highest recall (88.5%) in detecting 26 known vulnerabilities across 13 tested AI models.
• More expensive models are not necessarily better; cheaper variants or pooling results from mid-tier models can outperform a single run of a flagship model.
• Model runs are inconsistent, but running a model multiple times and pooling the findings (pass@3) significantly improves detection rates.
• Open-weight models like GLM-5.2 are demonstrating competitive performance, reaching 59% detection.

• Claude Mythos 5 leads BenchLM's July 2026 AI model rankings with a score of 83.9, followed by Claude Fable 5 (83.7) and GPT-5.6 Sol (82).
• The rankings are based on the BenchAlign v5 contract, with provisional scores calculated from 8 benchmark categories.
• Claude Fable 5 is identified as the top performer overall, with GPT-5.4 and Claude Opus 4.6 also close in scoring.
• The rankings consider both open-weight and proprietary models, but deployment terms and costs are not factored into the BenchAlign score.

• The LLM Leaderboard 2026 displays public benchmark performance for models released after April 2024, sourced from providers and independent evaluations.
• It features results from non-saturated benchmarks, excluding outdated ones like MMLU.
• "Humanity's Last Exam" shows a 70% score for the top overall model.
• Top models are also ranked for Reasoning (GPQA Diamond), Agentic Coding (SWE Bench), Work Automations (AutoBench), Computer Use (OSWorld), Browsing (BrowseComp), and Terminal Use (Terminal-Bench 2.1).
• The leaderboard also lists the fastest, lowest latency, and cheapest models based on token processing and cost per million tokens.

• Claude Fable 5 leads as the best overall AI model in July 2026, excelling in real-world coding correctness with 95% on SWE-bench Verified.
• The AI market in 2026 is characterized by numerous models with similar benchmark scores, making cost per finished task the primary differentiator.
• Key models for specific tasks include Claude Sonnet 5 for everyday coding, Gemini 3.1 Pro for reasoning and math, and Meta Muse Spark 1.1 for agents.
• Value models like Grok 4.5 and DeepSeek V4 offer competitive performance at significantly lower costs, making them attractive for budget-conscious applications.
• The recommended strategy for AI implementation involves using a tiered approach, routing routine tasks to cheaper models and reserving premium models for complex, correctness-critical tasks.

• SAP announced on July 17, 2026, that it has acquired Freiburg-based Prior Labs, a pioneer in Tabular Foundation Models.
• The company plans to invest over €1 billion during the next four years to operate Prior Labs as an independent AI research lab focused on structured business data.
• Prior Labs, founded by Frank Hutter, Noah Hollmann, and Sauraj Gambhir, will be advised by Yann LeCun and Bernhard Schoelkopf.
• This initiative aims to leverage structured data for enterprise AI, distinct from chatbot-focused advancements.

• Starting July 20, Anthropic is capping Claude Fable 5 access for Max and Team Premium users at 50% of their plan limits.
• On the same date, subscription limits for these plans will also be reduced by approximately one-third.
• Pro and Team Standard subscribers will receive a one-time $100 usage credit for Fable 5, after which they will be charged API rates.
• Anthropic cited demand management and investment in capacity as reasons for the changes, with Fable 5 being moved to higher-tier plans.

• AI lab DeepSeek secured $7.4 billion in its first external funding round, valuing the company at over $50 billion.
• China's National AI Industry Investment Fund received direct investment with voting rights and no lock-up period.
• Commercial investors, including Tencent, CATL, and JD.com, were placed in a limited partnership managed by the founder, with a five-year lock-up and no governance voice.
• This deal structure suggests Beijing views frontier AI model developers as national assets, potentially impacting international AI competition.

• Artificial intelligence company Anthropic will invest $50 billion to build AI data centers across the United States.
• The first data centers are planned for Texas and New York, with additional locations to follow.
• These facilities are expected to be operational throughout 2026 and will create 800 jobs.
• Anthropic states this investment is necessary to meet the growing demand for its AI chatbot Claude and to advance AI research.

• The Sony Bravia 9 II is a new flagship RGB LED TV featuring enhanced brightness, more dimming zones, and an anti-reflective screen compared to previous models.
• It offers excellent color accuracy and brightness suitable for well-lit rooms, making it a strong contender for bright-room viewing.
• Despite its advanced LED technology, the TV can exhibit some blooming, particularly when viewed off-axis.
• The Bravia 9 II is priced higher than comparable OLED models and the company's own Bravia 7 II, leading to a recommendation for OLEDs for most users seeking superior contrast.

• Moonshot AI launched Kimi K3, a 2.8 trillion parameter AI model, positioning it as their most capable model to date.
• The model is available via their website and API, with an open-weight release planned for July 27, 2026.
• Self-reported benchmarks suggest Kimi K3 performs competitively with leading models like Claude Opus and GPT-5.5, while also excelling in specific areas like frontend code generation.
• Kimi K3 introduces a higher pricing structure, making it the most expensive model released by a Chinese AI lab, though it shows improvements in token efficiency compared to previous versions.

• Fable 5 significantly outperformed GPT-5.6 Sol on an NP-hard optimization problem, achieving the best overall solution and exhibiting exceptional consistency.
• The article found that the "/goal" mode, designed to improve performance, did not consistently benefit either model and could even lead to worse outcomes on this specific task.
• While "/goal" won more individual trials, it negatively impacted the average performance of both Fable 5 and GPT-5.6 Sol.
• The study highlighted Fable 5's superior raw intelligence and consistency compared to GPT-5.6 Sol.

• A couple is considering purchasing their in-laws' $800,000 home for $400,000.
• Mortgage experts advise against simply lowering the purchase price.
• The recommended approach is to structure the transaction at the property's full market value, using a gift of equity and seller concessions to cover the buyer's down payment and closing costs.
• This method is suggested to provide a better loan-to-value ratio, facilitate qualification for a more favorable mortgage, and avoid out-of-pocket closing costs for the buyers.

• Rocket Lab is expanding its launch infrastructure, including ground systems, mission control, and launch sites.
• This expansion is intended to support more frequent missions and a wider range of customers, including commercial and national security clients.
• The company aims to use the enhanced facilities for more complex missions and to provide quicker access to space.

• SpaceX stock closed at a low Thursday night after the company scrubbed its planned Starship V3 test flight due to engine issues.
• CEO Elon Musk cited a failure of some engines to ignite, triggering an automatic abort, and indicated another attempt might occur in a few days.
• The aborted launch was SpaceX's 13th Starship test flight and the first since its IPO on June 12.
• The mission aimed to test booster separation, return, and Starship's ability to deploy satellites and perform a controlled descent.

• Google DeepMind has launched a three-month accelerator program for AI-focused startups, research teams, and non-profits in the Asia Pacific region that are addressing environmental challenges.
• The program, running from September to December 2026, will select 10-15 organizations to receive access to Google's AI stack and expert support.
• Participants will engage in an in-person bootcamp, followed by virtual mentorship and technical guidance, culminating in an in-person Demo Day.
• Applicants need a functional prototype, an AI-centric solution, an established technical team, and focus areas such as nature, climate, agriculture, sustainability, or energy.

• In June 2026, the data center hardware landscape saw significant developments across networking, memory, CPUs, orchestration software, and chip design, with vendors investing beyond GPUs to improve AI infrastructure utilization and reduce bottlenecks.
• Key announcements included HPE's AI infrastructure strategy, warnings about memory supply strains due to AI buildouts, and Nvidia's increased market share in data center Ethernet switching.
• Nvidia also detailed progress on its Vera Rubin platform and Vera CPU, alongside launching the DSX OS for AI factory management.
• IBM is exploring sub-1-nm chip architectures with its NanoStack initiative, while AWS introduced Graviton5-powered EC2 instances, underscoring the role of CPUs in AI.
• Other highlights included QumulusAI's focus on infrastructure utilization and Qualcomm's expansion into hyperscale infrastructure with a Meta CPU deal and new AI data center platform.

• Google's new STATIC framework enables significantly faster generative retrieval by using vectorized sparse matrices instead of prefix trees.
• This technology allows Google to enforce millions of constraints, such as content freshness or product availability, simultaneously with $O(1)$ I/O complexity.
• STATIC shifts search from keyword matching to Semantic ID (SID) optimization, requiring SEOs to focus on consistent entity data and structured markup.
• The framework's efficiency is expected to result in a higher consumption of fresh content and more precise, constrained AI answers in search results.

• A GitHub repository named static-search-tree offers two Rust crates: one to optimize binary searches in sorted integer arrays and another for faster suffix array searching.
• The integer array optimization provides a data structure that answers queries significantly faster with a 6% space overhead.
• The project is currently a research endeavor with code not intended for direct use as a dependency in other projects, and documentation is limited.
• Instructions are provided to reproduce experiments and generate plots, including notes on enabling SIMD instructions for compilation.

• New implicit data structures, named S-tree and S+ tree, have been developed based on static B-tree and B+ tree memory layouts.
• These structures accelerate searches in sorted arrays by leveraging SIMD instructions.
• The S-tree is up to 8x faster than std::lower_bound while using the same memory.
• The S+ tree is up to 15x faster, using slightly more memory.
• The optimizations are demonstrated using Clang 10 targeting a Zen 2 CPU, with performance expected to transfer to other platforms.

• Dataland, the world's first AI art museum, opened last week in Los Angeles, featuring immersive exhibitions.
• Founded by Refik Anadol and Efsun Erkılıç, the museum's inaugural exhibition, "Machine Dreams: Rainforest," utilizes an AI system trained on ecological data to create a multi-sensory experience.
• Visitors can interact with the exhibition through wearable biosensors and scent devices, influencing the evolving artwork with their biometric data and presence.
• The museum also includes interactive galleries for data exploration and a culinary program that translates ecological data into flavors, challenging traditional notions of art and audience engagement.

• In 2026, numerous new museums are set to open globally, including the Guggenheim Abu Dhabi, which will be the largest Guggenheim worldwide and focus on post-1960s art from West Asia, North Africa, and South Asia.
• Other notable openings include the Lucas Museum of Narrative Art in Los Angeles, dedicated to illustration, cinema, and popular culture, and the V&A East Museum in London, which will embed national collections within a diverse community.
• The Obama Presidential Center in Chicago will function as a community center and museum, departing from traditional presidential library models.
• These institutions collectively aim to redefine the role of museums, acting as platforms for narrative, identity, and cultural agency in public spaces.

• Los Angeles will see multiple significant museum openings in 2026, including the $720m expansion of LACMA in April and the $1bn Lucas Museum of Narrative Art on September 22.
• Refik Anadol's Dataland, the first museum dedicated to AI-generated art, is also set to open in downtown Los Angeles in the spring.
• In London, V&A East will open on April 18, aiming to showcase applied and decorative arts collections to new audiences.
• Crystal Bridges Museum of American Art in Arkansas will expand its facilities, opening on June 6, while the Mosul Museum in Iraq is scheduled to reopen in November after extensive reconstruction.
• The London Museum will relocate to a new £437m site in late 2026, and the Guggenheim Abu Dhabi is also expected to open its doors that year.

• The Ministry of Awe, an immersive art experience in Philadelphia, opens on March 14, 2026, in a former 19th-century bank.
• The Centre for Contemporary Arts Tashkent, Uzbekistan's first permanent institution dedicated to contemporary art, opens on March 21, 2026.
• InnovateX, a forum exploring the future of experiences, takes place on March 10, 2026, at the Science Museum in London.
• Inside Aardman, an exhibition celebrating 50 years of the animation studio, is open now until November 15, 2026, in London.

• The world's first museum dedicated to A.I.-generated art, Dataland, is opening in Los Angeles on June 20.
• The museum features immersive, multisensory exhibitions created by artificial intelligence, with its inaugural show focusing on rainforests.
• Dataland's creators state they source their data ethically and utilize renewable energy for their cloud servers to address concerns about copyright and environmental sustainability.
• While some art critics express philosophical opposition to A.I. art, public reception appears to be more favorable.

• Banks are leveraging artificial intelligence (AI) to significantly transform financial services, improving fraud detection, customer service, and operational efficiency.
• AI applications in banking include enhancing credit risk analysis, regulatory compliance, investment decision-making, and automating operational processes.
• Emerging agentic AI allows for autonomous agents to manage complex workflows, while strong data governance and ethical AI frameworks are crucial for trust and compliance.
• McKinsey estimates that strategic AI use could add $1 trillion annually to the banking sector's value.

• Artificial intelligence (AI) is significantly transforming the banking industry by improving efficiency and driving revenue growth.
• Banks that fully embrace AI could see up to a 15-percentage-point improvement in their efficiency ratio, acting as a barometer for AI maturity.
• AI enables banks to personalize customer experiences, automate routine tasks, and reallocate talent to higher-value activities, leading to smarter growth and operational agility.
• These advancements require substantial investments in infrastructure, data governance, and responsible AI oversight.

• Apple is reportedly skipping the M6 Pro, Max, and Ultra chips to prioritize AI development with the M7 generation.
• The first M7 Macs are expected in early 2027, with higher-end M7 Pro and Max models later that year, and M7 Ultra in 2028.
• This strategy shift focuses on significant Neural Engine upgrades and increased memory support, with the M7 Ultra targeting up to 1.5 terabytes of unified memory.
• This AI-centric chip development extends to Apple's server strategy, with M7 Ultra-derived chips planned for next-generation AI servers around 2029.

• Apple's semiconductor development from 2015-2026 is categorized into three phases: Foundation, Neural Acceleration, and Unified Architecture Dominance, involving over $20 billion in investment and significant team growth.
• The Neural Engine's performance dramatically increased from 0.6 TOPS in 2017 (A11 Bionic) to 38 TOPS in 2024 (M4), enabling on-device large language models.
• The M1 chip in 2020 marked a shift to Apple Silicon for Macs, featuring a unified memory system for improved performance and efficiency.
• Apple has strategically protected its custom silicon through over 29 Neural Engine patents filed between 2018 and 2025, focusing on core operations and architecture.

• Apple is reportedly shifting its M-series Mac chip roadmap to prioritize artificial intelligence (AI) capabilities, with future chips like the M7 family designed for significant AI performance gains.
• Technologies developed for Apple's canceled self-driving car project, which reportedly cost over $10 billion, have formed the foundation for its AI hardware strategy and the Neural Engine.
• The M7 chip family, expected to begin arriving in late 2026 or early 2027, will reportedly skip Pro, Max, and Ultra variants of the M6 generation to accelerate AI advancements.
• The upcoming M7 Ultra chip is anticipated to offer substantial AI performance increases, potentially rivaling dedicated AI accelerators, and will support up to 1.5TB of memory.
• Development of the M8 chip generation, also focused on enhanced AI capabilities, is already underway, with chips expected in 2028.

• Dataland, the world's first "museum of AI arts," opened in downtown Los Angeles on June 20, co-founded by artist Refik Anadol.
• The inaugural exhibit, "Machine Dreams: Rainforest," uses AI to create immersive, interactive experiences that respond to visitor movements and biometric data.
• The exhibit draws from Anadol's AI models trained on vast datasets of natural science archives, aiming to redefine AI art beyond simple prompt generation.
• Dataland emphasizes ethical data sourcing, environmental responsibility, and explores the concept of artwork "feeling" visitors back through sensory and biometric feedback.

• The article analyzes Joel Yliluoma’s 2011 ordered dither algorithm, providing an in-depth explanation of its mechanics.
• It introduces simplified variants of the algorithm and compares their results to a state-of-the-art method.
• The discussion covers both black and white and color image dithering, detailing techniques for color selection and candidate color generation.
• The article draws parallels with Thomas Knoll's dither algorithm, famously used in Adobe Photoshop, and explores a geometric interpretation of Yliluoma's approach involving exponential moving averages.

• A static search tree (S+ tree) has been developed that is up to 40 times faster than binary search for querying sorted data.
• The optimizations focus on improving throughput by processing queries in batches and leveraging CPU architecture, including SIMD instructions and efficient memory layout.
• The S+ tree reorders data to better utilize CPU cache lines, fetching multiple data points at once to reduce memory access latency.
• This development is motivated by the need for faster indexing in fields like bioinformatics, specifically for speeding up suffix array searches.

• In 1988, a Scottish hi-fi company named Linn Products developed a custom computer called Rekursiv, which featured hardware memory safety, automatic garbage collection, and treated memory and disk as a single persistent object store.
• Despite its advanced concepts, the Rekursiv was commercially unsuccessful due to the rapid advancements and lower costs of commodity microprocessors at the time.
• The computer's hardware and associated media were reportedly discarded in the Forth and Clyde Canal.
• Rekursiv's design principles, particularly in hardware memory safety, are now being implemented in modern silicon from companies like Arm.

• Wall Street's five largest banks collectively reported $114 billion in capital markets revenue in the first six months of 2026, a 31.5% increase from the previous year, largely driven by stock trading, dealmaking, and financing related to the AI boom.
• Analysts cite AI as the primary earnings driver for these banks in 2026, with increased capital demands from AI-focused tech and related industries fueling lucrative Wall Street activities.
• This surge includes significant contributions from stock trading, wealth management benefiting from new AI wealth, and advisory fees from AI-related IPOs and equity raises.
• While optimistic about the current banking environment, executives acknowledge the dependence on the AI revolution and anticipate a multi-year AI build-out cycle, with significant investment yet to occur.

• Apple is reportedly accelerating its M-series chip release roadmap, prioritizing AI performance for its upcoming M7 generation processors.
• This acceleration may involve skipping certain processor variants, like Pro, Max, and Ultra versions of the M6 chip expected this fall, to focus on developing advanced AI capabilities.
• The company is aiming to enhance on-device AI processing power, potentially pushing local performance to workstation class by late 2027.

• Apple's M6 chip roadmap is reportedly being influenced by the AI boom, potentially leading to a shift in its silicon development strategy.
• Instead of fully developing the M6 lineup, Apple may accelerate the rollout of the M7 chip, with M7 Pro and Max variants expected in late 2027.
• The M7 Ultra, anticipated in 2028, could feature significant neural processing upgrades for AI performance and potentially support up to 1.5TB of memory.
• This strategic acceleration aims to enhance Apple's AI server infrastructure and compete with rivals investing in dedicated AI hardware.

• Apple's M-series chip roadmap is shifting to prioritize AI performance, with the M6 expected in late 2026, featuring only a base chip.
• The company will skip M6 Pro, Max, and Ultra variants, proceeding directly to the M7 family in 2027, which will include Pro, Max, and Ultra versions by 2028.
• The M7 Ultra is designed to support up to 1.5TB of memory and is intended for a future Apple AI server.
• The M8 generation is planned for 2028, utilizing a 1.4nm manufacturing process for further performance and efficiency gains in on-device AI.

• Apple has released the Game Porting Toolkit 4, featuring AI coding agent support to streamline bringing games to Apple platforms.
• Updates to Metal 4 and MetalFX improve visual quality and developer workflow, with a redesigned temporal upscaler leveraging M5 Pro and M5 Max hardware.
• New StoreKit plug-ins for Unity and a Steam Asset Converter are available for games on the App Store, alongside improvements for faster game loading.

• By 2026, the building management market is shifting from data monitoring to autonomous operations driven by AI that triages issues and initiates actions.
• This transition prioritizes reducing human decision-making by enabling AI to explain, recommend, and document solutions, moving away from an "alarm economy."
• Wireless Building Management Systems (BMS) are becoming a standard starting point for retrofitting, focusing on establishing a reliable data layer before implementing targeted controls.
• The emphasis for building teams is on fewer comfort complaints, reduced reactive maintenance, and predictable energy costs, achieved by AI that quietly removes operational friction.

• The AI in smart buildings and infrastructure market is projected to grow by 2026, with further forecasts extending to 2035.
• The market is segmented by type (software, hardware, services), technology (machine learning, NLP, computer vision, RPA, etc.), and application (building automation, energy management, security, maintenance, etc.).
• Key end-user industries include commercial, residential, industrial, government, healthcare, educational, and retail sectors.
• The report analyzes market size, trends, strategies, and provides forecasts.

• The AI in smart buildings and infrastructure market is projected to reach $93.48 billion by 2030, with a compound annual growth rate (CAGR) of 21.1%.
• Growth drivers include the expansion of smart city initiatives, increasing sustainability mandates, advancements in edge AI, demand for automated facility management, and integration with digital twins.
• Key trends encompass AI-driven building automation, predictive maintenance, energy optimization, intelligent security, and occupant-centric systems.
• North America held the largest market share in 2025, and tariffs are impacting project deployment by increasing costs for imported hardware.

• Apple has released the Game Porting Toolkit 4, a toolset designed to simplify bringing games to Mac, iPad, and iPhone.
• The toolkit allows developers to assess how Windows games will perform on Mac, convert graphics and shaders, and leverage Apple silicon's capabilities.
• New agent skills and command-line Metal tools are included to allow AI coding agents to manage the porting process, potentially reducing time and cost.
• Apple also provides hardware, graphics, audio, social gaming, and distribution tools to support game development across its devices.

• Apple has released the Game Porting Toolkit 4, designed to streamline bringing games to Mac, iPad, and iPhone.
• The toolkit integrates with AI coding agents via a GitHub repository for improved efficiency and quality in game porting.
• It now supports Metal 4 for compatibility testing and allows direct debugging of Metal workloads using command-line access.
• Developers can evaluate Windows binaries on Apple silicon to assess performance and shader compatibility before porting.

• AI hardware startup Aina has secured $5.5 million in seed funding, led by Redstart Labs and 360 ONE Asset.
• The company plans to use the capital to commercialize its flagship interface and expand its teams in San Francisco and Bengaluru.
• Aina previously launched Dune, a context-aware keypad for Mac, and has shipped units to early adopters.

Here's a summary of the article:

• Siemens' blog post explores the transformative impact of Artificial Intelligence (AI) on smart building management.
• AI integration enables predictive maintenance, optimizes energy consumption, and enhances occupant comfort and safety.
• The technology allows for real-time data analysis, leading to more efficient building operations and reduced costs.
• AI in smart buildings facilitates proactive issue resolution and a more personalized environment for users.

• Smart building automation is being transformed by the integration of IoT and AI, enabling buildings to operate with increased autonomy and precision.
• By 2025, AI and IoT-powered buildings are projected to achieve significant reductions in energy consumption (up to 30%) and maintenance costs (25-30%), while increasing equipment uptime by over 70%.
• The global smart building market is expected to grow substantially, driven by factors such as energy regulations, ESG mandates, rising energy costs, and tenant demand for improved indoor environments.
• IoT sensors provide the necessary data for AI systems, allowing for predictive maintenance, adaptive energy optimization, and digital twin simulations, shifting building operations from reactive to proactive.

• Smart buildings are evolving into "live systems" that use AI to predict, adapt, and self-correct, moving beyond traditional construction materials.
• These advanced buildings are equipped with embedded sensors and algorithms to manage environmental factors like temperature, lighting, water usage, and carbon intensity in real time.
• AI enables predictive maintenance and real-time optimization of energy consumption, potentially reducing energy footprints by 8-19% over three decades.
• Smart building systems are shifting from reactive problem-solving to proactive anticipation, improving efficiency, comfort, and operational costs.

• Apple silicon processors offer significant performance improvements for gaming, with faster processing speeds, smoother graphics, and reduced load times compared to previous Intel chips.
• The unified memory architecture optimizes graphics-intensive tasks, leading to visually enhanced gaming experiences and higher frame rates.
• The availability of games on macOS is increasing, with more developers porting titles and utilizing tools like Apple's Metal API, and platforms like Steam support many games.
• Software like Crossover allows users to run Windows games on Apple silicon devices, and cloud gaming services provide further options for accessing a wider game library.

• Apple is expected to launch its next-generation M6 chip later this year, featuring upgrades to memory bandwidth, core performance, and a redesigned GPU.
• The M6 chip will reportedly increase memory bandwidth to 200 GB/s, boosting on-device AI performance, and will feature up to 12 graphics cores.
• Unlike previous generations, Apple is reportedly planning to release only the base M6 processor and no higher-end Pro, Max, or Ultra variants.
• This change in strategy is attributed to anticipated significant improvements in the subsequent M7 chip, expected in early 2027.
• The M6 chip has been tested in an updated base model MacBook Pro and may be the only product to feature this generation of Apple Silicon.

• Apple's M5 chip, featuring Neural Accelerators and the Metal 4 API, has demonstrated a significant increase in gaming performance on Macs.
• A user on an M5 Max MacBook Pro achieved 60-70 FPS in Resident Evil Requiem (emulated via CrossOver) at QHD resolution with Frame Generation enabled, without upscaling.
• This advancement suggests a strategic shift towards AI for gaming performance on Apple Silicon, moving beyond traditional rasterization.
• Potential latency issues with Frame Generation were noted, but the overall performance boost highlights the capabilities of Apple's Neural Accelerators for future gaming adoption.

• Funding for foundational AI startups doubled in Q1 2026 compared to all of 2025, with OpenAI, Anthropic, and xAI raising over $160 billion combined in the past year.
• Top venture capital firms investing in AI include Khosla Ventures, AI Fund, AIX Ventures, Conviction, and Gradient Ventures, with Andreessen Horowitz, Sequoia, Founders Fund, and Coatue leading foundational-model rounds.
• While large foundation-model deals absorb significant capital, there is ample funding available for application-layer AI, vertical AI, and AI infrastructure startups.
• Successful pitches in 2026 emphasize proprietary data or a strong vertical focus, rather than generic LLM wrappers.

• In Q2 2026, 1,516 companies raised $150 billion, with debt financing accounting for 63.9% of disclosed capital.
• June was the highest funding month of the year, raising $83.1 billion.
• AI companies comprised 35.8% of funded companies and captured 49.2% of disclosed capital.
• Early-stage funding (pre-seed to Series A) represented 46.4% of deals but only 5.1% of capital.
• California saw its share of disclosed dollars decrease from 63.7% to 44.1% as debt financing flowed to projects in other states.

• Yutori, an AI startup, has announced it has raised $4.6 million in seed funding.
• The company was founded by former Google and Stripe employees.
• Yutori develops AI models and tools for developers.
• The funding round was led by Gradient Ventures.

• Apple and Major League Baseball have entered into a multi-year agreement to equip every MLB team with iPad Pro tablets.
• The 12.9-inch devices, housed in rugged, MLB-branded cases, will feature a custom app named MLB Dugout.
• This app will provide managers with access to performance statistics, game videos, and analytical data on pitcher-hitter matchups.
• The deal lifts a previous ban on laptops, smartphones, and tablets in MLB dugouts and allows announcers to refer to the devices as "iPads."

• AI is transforming building management in the US and Canada, offering smarter, more efficient, and autonomous systems.
• AI-powered systems can reduce energy consumption by up to 30% in commercial buildings and are projected to grow significantly.
• Predictive maintenance using AI can reduce maintenance costs by 10-15% and extend asset lifespans.
• AI integration helps buildings meet stricter environmental regulations and Environmental, Social, and Governance (ESG) goals.

• A conceptual framework and prototype assessment were developed for LLM-based AI agents to manage energy in smart buildings.
• The system uses perception, central control, and action modules for context-aware energy management and natural language interaction.
• Evaluations on real-world residential energy datasets showed promising accuracy for device control (86%) and memory tasks (97%), but lower accuracy for cost estimation (49%).
• The study aims to formalize assessment for these AI agents and identify future research directions.

• Build has partnered with Index. to accelerate institutional development using AI agents.
• AI agents in real estate are autonomous systems that execute commercial real estate workflows independently, going beyond chatbots to act as virtual team members.
• These agents can perform tasks such as site sourcing, due diligence, underwriting, market research, and reporting, creating institutional-grade deliverables.
• The "agentic development" model involves AI agents working alongside human experts to execute the full development lifecycle, from site selection to investment preparation.

• Apple Silicon Macs, particularly those with M3 Max and M2 Ultra chips, demonstrate improved gaming capabilities, running demanding titles like "Shadow of the Tomb Raider" at playable frame rates.
• While base M1, M2, and M3 chips are capable, higher-tier models like M3 Pro, M3 Max, and M2 Ultra offer a significantly superior gaming experience.
• The primary limitation for gaming on Macs remains the limited availability of native macOS titles, though Apple's new game porting tool shows promise for future game development.
• Streaming services like Steam Link and cloud gaming platforms such as Nvidia GeForce Now, along with Apple Arcade's curated selection, offer alternative ways to play games on Mac.
• Support for popular game controllers like Sony's DualSense and Xbox controllers enhances the gaming experience on macOS.

• New tests indicate Windows games are performing significantly better on Macs, both officially and unofficially.
• Apple's Game Porting Toolkit, designed to assist developers in porting games, is also being used by users to run Windows software.
• The latest beta version of the toolkit shows substantial performance increases, with Grand Theft Auto V seeing an approximate 66% frame rate boost in recent tests.
• These improvements are hoped to encourage game studios to consider Macs as a viable platform for their titles.

• Apple's Game Porting Toolkit 4 beta has significantly improved Mac gaming performance, with Grand Theft Auto V seeing a 66% frame rate increase on an M4 Pro MacBook Pro.
• The toolkit allows developers to test and run Windows games on macOS by translating DirectX to Apple's Metal graphics API.
• This advancement enhances compatibility and leverages Apple Silicon's gaming capabilities, making Macs a more viable gaming platform.

• AI agent startups secured over $1.8 billion in July 2026 across more than 12 deals, a 35% increase from June.
• Enterprise automation agents received the largest share of funding, accounting for 58% of the total, indicating investor preference for B2B monetization.
• Developer tooling agents also saw significant investment, raising $420 million with strong market validation in coding assistance.
• Sequoia Capital was the most active investor, leading four deals, and average valuations for AI agent startups climbed 40% quarter-over-quarter to $280 million.

• Recent funding rounds are being raised by various AI startups, including those focused on full-stack AI cloud infrastructure, embodied AI via home cleaning services, and AI-powered vehicle inspection.
• Other companies securing capital are developing AI for personalized wedding planning, AI agents, and AI-native operations for insurance firms.
• Innovations span AI-powered drug discovery, robotic sculpting factories, and AI infrastructure deployment.
• Funding is also directed towards AI for quality control in manufacturing and drone defense systems.

• JetBrains has released the Release Candidate (RC) for Rider 2026.2, an integrated development environment for .NET and game developers.
• Key new features include native GitHub Copilot integration, enhanced AI coding agent capabilities using the IDE's intelligence, and significant performance improvements for debugger launches and branch switching.
• The release also offers faster Unreal Engine project indexing, WPF Hot Reload, a redesigned NuGet tool window, and expanded support for Godot and Azure development.

• Organizations increasingly use AI, but scaling it across enterprises remains a challenge.
• Leading AI development platforms include GitHub Copilot for coding assistance, Gemini Enterprise Agent Platform and Microsoft Foundry for cloud AI, Zapier for workflow automation, and OutSystems for enterprise-grade agentic development.
• Platforms vary in scope, from coding assistants to comprehensive enterprise solutions supporting the full AI application lifecycle.
• Choosing the right platform depends on specific needs, such as speed, developer control, model development, or enterprise-scale governance and integration.

• AI chip startup Etched is reportedly in discussions for a funding round at a $20 billion valuation.
• A separate funding round, led by Sequoia, is reportedly being raised at a $10 billion valuation.
• The company recently exited stealth at a $5 billion valuation and has disclosed $800 million raised to date.
• Etched's transformer-only inference chip, Sohu, is designed for a specific AI workload.

• Valar Atomics, an El Segundo, California-based startup, is reportedly in talks to raise $1 billion in equity at a valuation of approximately $6 billion, with Sequoia expected to lead the funding round.
• This potential valuation follows a recent demonstration where Valar's small modular reactor successfully powered an Nvidia AI chip, leading to a data-center partnership announcement.
• Previously, the three-year-old company raised $450 million in March 2026 at a $2 billion valuation.

• A 15-year-old ASUS Eee PC 1000HE netbook with an Intel Atom N280 processor and 1GB of RAM was revived by installing Arch Linux 32.
• The author chose Arch Linux 32 due to the netbook's 32-bit architecture and limited hardware resources, requiring a lightweight operating system.
• The installation process involved preparing boot media, connecting to the internet via Wi-Fi, setting the system clock, and partitioning the netbook's hard drive.
• The author created a 4GB swap partition and a primary Linux partition for the Arch Linux installation.

• Apple's Game Porting Toolkit 4 beta shows significant performance gains for Mac gaming, with Grand Theft Auto V experiencing a 66% frame rate increase on an M4 Pro MacBook Pro.
• The toolkit translates Windows DirectX calls to Apple's Metal API in real-time, enabling Windows-exclusive games to run on Macs and aiding developers in evaluating ports.
• These software advancements suggest Apple Silicon's hardware is no longer the primary limitation for Mac gaming performance.

• Dev Technosys, Google Cloud Vertex AI, Microsoft Azure AI, IBM watsonx, and Amazon Web Services (AWS) AI are highlighted as the top 5 artificial intelligence development platforms in 2026.
• These platforms offer end-to-end ecosystems for building, deploying, and scaling AI applications, integrating with cloud infrastructure and enterprise systems.
• Key features include machine learning development, generative AI, AI agents, computer vision, natural language processing, and enterprise automation.
• Choosing the right platform depends on defining business objectives, evaluating scalability, considering integration capabilities, prioritizing security, reviewing development tools, and analyzing total cost.

• In 2026, AI development is characterized by a vast array of tools and a shift towards using pre-trained models accessible via APIs or downloadable from hubs.
• AI development tooling has evolved from assistive to augmentative, and now to agentic tools capable of autonomous multi-step tasks.
• Three main audiences build production-grade AI applications: ML framework users, AI coders using assistants, and non-engineers using prompt-to-app builders.
• Key tool categories include ML frameworks (e.g., PyTorch), model hubs and APIs (e.g., Hugging Face), LLM development frameworks, AI coding assistants, deployment and serving tools, and environment management.
• PyTorch is identified as the best ML framework due to its flexibility and extensive ecosystem, while Hugging Face is the leading model hub and API service for accessing pre-trained models.

• Siemens has launched the Eigen Engineering Agent, an AI-powered tool designed to bring artificial intelligence into the physical world.
• This new agent integrates AI capabilities into physical systems, enabling them to understand and interact with their environment.
• The technology aims to enhance automation and decision-making processes in various industrial applications.

• Siemens has introduced AI agents designed for industrial automation.
• These agents are intended to enhance operational efficiency and flexibility within industrial settings.
• The introduction marks a step towards integrating advanced AI capabilities into manufacturing processes.

• Human-centric autonomous buildings analyze, learn, predict, and act to optimize indoor experiences and operational efficiency.
• These buildings go beyond smart systems by using AI agents for self-optimization and self-healing, minimizing intervention.
• They aim to enhance occupant well-being and productivity, while providing operators with strategic decision-making focus and owners with cost and energy efficiencies.
• The goal is to create healthier, more productive environments by anticipating occupant needs through learned behavioral patterns.

• Siemens offers Industrial AI solutions aimed at transforming companies into Digital Enterprises.
• These solutions facilitate a continuous data flow from design through realization and optimization.
• A unified data fabric provides contextualized data across the enterprise, enabling cross-domain intelligence.
• Siemens' AI solutions aim to help businesses convert data into confident actions throughout their value chains.

• California lawmakers are considering a series of bills in the 2025-2026 session to regulate the use of artificial intelligence, automation, and surveillance in the workplace, with major unions supporting these measures.
• Proposed legislation includes limitations on AI in employment decisions, restrictions on invasive AI surveillance, and requirements for transparency regarding AI tool usage and data collection.
• Several bills aim to protect jobs and human roles in sectors like healthcare and higher education, and to mandate advance notice for technological displacement.
• These bills, if enacted, could significantly increase compliance burdens for employers and introduce new legal uncertainties regarding workplace technologies.

• New York City's Local Law 144 requires annual independent bias audits for automated employment decision tools (AEDTs) used to evaluate NYC residents for employment or promotion.
• The law also mandates notification to candidates and employees about AEDT use and public transparency regarding audit results.
• AEDTs are defined as computational processes using machine learning, AI, or similar technologies that significantly assist or replace discretionary decision-making.
• Compliance involves conducting audits, providing notice, and publishing a summary of audit findings, with penalties for violations.

• NYC Local Law 144 requires employers using automated tools to screen, score, or select candidates for jobs based in New York City to conduct annual independent bias audits.
• Compliance applies regardless of the employer's headquarters location, extending to remote roles filled by NYC residents.
• Employers must notify candidates when an automated tool is used and publicly disclose the summary of the bias audit on their website.
• Failure to comply can result in daily financial penalties and increased risk of discrimination claims, impacting the employer's brand.

• New York City's Local Law 144 (LL 144), enacted in July 2023, requires employers using automated employment decision tools (AEDTs) for hiring or promotion for jobs in NYC to conduct annual independent bias audits for race and gender.
• Employers must publicly post summaries of these audits, including start and audit dates, data sources, applicant counts, and key metrics, for at least six months.
• Jobseekers must be notified at least ten business days before an AEDT is used in the hiring process and be given the option to opt out of automated screening.
• LL 144 applies to any employer, agency, or recruiter using AEDTs for jobs performed in or associated with NYC, regardless of the company's physical location.
• The law is enforced by the NYC Department of Consumer and Worker Protection, with penalties for violations.

• Alpaca, a provider of brokerage infrastructure, has raised $135 million, bringing its total financing package to $435 million with debt.
• The funding, led by Peak XV and including Elefund, Unbound, and Opera Tech Ventures, aims to support Alpaca's expansion in tokenized capital markets and AI-driven financial services.
• The company has recently expanded its regulated presence in India and Europe and has seen significant growth in API users and assets under custody for tokenized equities.
• Alpaca's infrastructure supports financial firms in building investing products across both traditional and on-chain markets.

• AI agents are transforming the engineering sector by automating complex workflows and accelerating product development.
• Unlike traditional automation, agentic AI can interpret intent, reason, adapt to deviations, and leverage context without explicit instructions.
• These agents eliminate manual, repetitive tasks, enabling engineers to focus on innovation and accelerating feedback cycles through rapid simulation setup and execution.
• By combining predictive Physics AI with agentic Engineering AI, SimScale enables rapid design exploration, optimizing hundreds of design variations in minutes.
• Trust in AI is built through transparency, allowing engineers to inspect every decision an agent makes, and the future involves collaborative, agent-to-agent workflows for a more integrated product development lifecycle.

• A new multi-agent framework integrates AI agents into automotive design to enhance creativity and efficiency.
• This framework automates tasks like sketching, 3D modeling, and aerodynamic simulations, reducing completion times significantly.
• It utilizes Vision-Language Models for generation, Large Language Models for orchestration, Geometric Deep Learning for 3D shape management, and AutoGen for agent coordination.
• The system comprises four agents: Styling for visual appeal, CAD for design retrieval and generation using DeepSDF, Meshing for CFD preparation, and Simulation for aerodynamic prediction.

• Researchers have introduced "Design Agents," an AI framework for engineering design, with a focus on automotive aesthetics and aerodynamics.
• This multi-agent system integrates with existing workflows to automate tasks like sketching, 3D modeling, and aerodynamic simulations, reducing design cycle times from weeks/days to minutes.
• The agents utilize advanced AI models, including VLMs and LLMs, and have been validated against industry benchmarks with high-fidelity simulations.
• The framework aims to accelerate design exploration and optimization for engineers and designers.

• The Layer Based SVG Engine (LBSE) has been developed for WebKit, aiming to integrate SVG rendering with the existing HTML/CSS machinery for performance gains.
• Early development involved creating a separate RenderLayer for every SVG element, which was found to be inefficient and memory-intensive.
• Recent work, supported by Wix and Igalia, has focused on reducing this overhead by implementing conditional layer creation, where layers are only generated when necessary.
• This change aims to improve performance, particularly in scenarios with numerous SVG elements and transformations, by reducing per-frame bookkeeping and memory usage.
• Further development is needed to optimize filter handling and address other complexities arising from the new conditional layer system.

• Siemens has launched the Fuse EDA AI Agent, an autonomous system to orchestrate complex electronic design automation (EDA) workflows.
• This agent integrates with existing tools for RTL coding, physical implementation, and manufacturing sign-off, automating multi-step tasks.
• It is designed to interpret EDA data, manage tool relationships in IC and PCB design, and supports third-party integrations.
• The system operates within secure HPC clusters, uses specialized RAG and multimodal data support to reduce AI hallucinations, and allows for human oversight in critical decisions.

• AI agents are emerging as intelligent collaborators for mechanical engineers, capable of understanding context, analyzing designs, and automating entire product development workflows.
• These agents go beyond traditional AI assistance by integrating directly into engineering processes, from CAD generation and simulation to design validation and change analysis.
• Notable AI agents include bananaz (for design validation and automated reviews), Siemens NX Design Copilot (for enterprise CAD), SOLIDWORKS AI Companions (for SOLIDWORKS users), Autodesk AI Assistant (for cloud-based development), PTC Creo Generative AI Tools (for generative design), and Claude Code (for coding and ideation).
• By automating repetitive tasks and catching issues early, AI agents aim to accelerate product development, reduce risk, and improve product quality, allowing engineers to focus on higher-value design thinking.

• AI agents are being evaluated and deployed in engineering design for manufactured products to improve consistency and reviewability.
• These agents execute multi-step engineering workflows autonomously, using structured data like CAD and drawings, and return outputs to engineers for review.
• Ideal workflows for AI agents involve tasks requiring high consistency, handling large amounts of information, high-volume execution, and retrieving scattered context.
• Current practical applications include CAD/drawing review, engineering search and decision context retrieval, agentic CAD tool operation, simulation setup assistance, and lessons learned knowledge management.

• As of July 2026, there are 607 AI infrastructure startups with verified founder contact information and fundraising data.
• These companies, building foundational technologies for AI applications, have collectively raised $332.0 million, with an average funding of $13.3 million per company.
• The startups are located across 7 countries, with San Francisco and New York being key hubs for this sector.

• AI infrastructure startup Fireworks has raised $1.5 billion in a Series D funding round, valuing the company at $17.5 billion.
• The funding round was led by Atreides Management, Index Ventures, and TCV, with participation from Nvidia Corp.
• Fireworks' valuation reflects its rapid sales growth, surpassing $1 billion in annualized revenue, and its customer base includes major tech firms like Samsung Electronics and GitLab.
• The company's platform allows developers to train and run AI models by fine-tuning open-source models, providing managed graphics card clusters and an AI agent to automate the workflow.

• Fireworks AI, a developer of enterprise AI tools, secured $1.5 billion in Series D funding, the largest of the week's reported funding rounds.
• The week saw significant investment in AI-focused startups, with other notable rounds including $650 million for meal and delivery provider Wonder and $400 million for AI drug discovery company Chai Discovery.
• Other companies receiving substantial funding include Walden Robotics for general-purpose robots and Brinc for drones, reflecting continued investment across various tech sectors.
• The funding period covered July 11-17, with a focus on U.S.-based companies.

• AI surveillance at work is leading to increased monitoring of employees, including automated firings and tracking of activities like bathroom breaks.
• In Europe, unions have been proactive in seeking protections against intrusive AI monitoring, with some German companies establishing models that prohibit algorithmic firings and limit the use of data for discipline.
• US labor unions are beginning to push for similar protections, focusing on transparency, notification of monitoring, and ensuring AI is used for training rather than solely for evaluation or discipline.
• Concerns include AI's potential for errors, discriminatory outcomes, and negative impacts on job quality and employee stress, though some employer groups acknowledge the need for dialogue and potential benefits.

• U.S. senators have introduced legislation, the Stop Spying Bosses Act and the No Robot Bosses Act, to regulate employers' use of AI, automated scoring systems, and surveillance technologies.
• The proposed bills would restrict the types of workplace data employers can collect and use for management, including biometric identifiers and behavioral inferences, and limit reliance on automated systems for consequential employment decisions.
• These regulations would apply to private employers with at least 11 workers and public agencies, covering both applicants and existing employees.
• Key provisions include prohibiting data collection for purposes like identifying union activity, requiring transparency in data usage, and allowing workers to opt out of automated decision-making processes.

• Meta paused its "Model Capability Initiative" (MCI) after over 1,600 employees signed a petition protesting a tool that tracked keystrokes, mouse clicks, and screen content for AI training.
• The initiative aimed to collect data from corporate laptops to improve AI models by observing how skilled employees work, a shift from traditional monitoring focused on security or productivity.
• Employees voiced concerns about their work behavior being converted into AI training material, including private conversations and developmental drafts, raising questions about data accessibility and consent.
• The article notes that such surveillance can alter employee behavior, potentially hindering the "messy" aspects of expertise that AI models learn best from.

• A new AI model developed by researchers at Chalmers University of Technology and the University of Gothenburg can simulate molecular evolution over 10,000 times faster than traditional methods.
• This AI model, presented in Science Advances, learns the dynamics of molecules and can predict their behavior and transitions on much longer timescales than previously possible.
• The technology has been tested on over 12,500 organic molecules and over a thousand short peptides, with results consistent with physics and previous studies.
• This breakthrough could significantly accelerate the costly and time-consuming early stages of drug discovery by identifying promising drug candidates more quickly and accurately.

• AI-driven drug discovery has attracted over $2 billion in investment, significantly reducing development timelines by 70% and doubling clinical success rates.
• These AI technologies compress traditional drug discovery phases from 4-5 years to 12-18 months, while also decreasing R&D costs by 30% to 40%.
• AI-validated targets show a 2.5 times greater probability of clinical progression, with regulatory approval rates increasing from 10-15% to 20%.
• The United States leads global investment in this sector, with major companies and pharmaceutical giants integrating AI into their research and development strategies.

• AI firm Databricks is reportedly close to a funding round that could value the company at $188 billion.
• The round is led by existing investor Coatue and will include both new and existing investors.
• Funds will be allocated to Databricks' multi-AI governance solution (Unity AI Gateway), AI coworker (Genie), and serverless Postgres database (Lakebase).
• The company also plans to invest in future AI acquisitions and deepen AI research.
• This funding follows a February round where Databricks raised $7 billion at a $134 billion valuation.

• Databricks is planning a strategic funding round that values the company at $188 billion.
• The raise, led by Coatue, is expected to finalize this summer and will focus on accelerating Databricks' AI strategy.
• Funds will be directed towards products like Unity AI Gateway, Genie, and Lakebase, designed to address challenges in enterprise AI deployment, cost management, and data governance.
• Databricks' platform aims to unify data and AI infrastructure for organizations across various industries, including finance.

• DrDroid, a Y Combinator-funded startup building AI agents for platform and infrastructure teams, is hiring a full-time Product Engineer in Bengaluru, India.
• The role requires over one year of experience in understanding product requirements and building solutions, with knowledge of distributed systems.
• Applicants must complete a mandatory assignment within seven days of submission, followed by interviews with the CTO and an engineer.
• DrDroid aims to help engineering teams automate triage, debugging, and remediation to fix issues faster.

• Isomorphic Labs has developed the Drug Design Engine (IsoDDE), a computational system that demonstrates significantly improved predictive accuracy over AlphaFold 3 for drug design.
• IsoDDE more than doubles AlphaFold 3's accuracy in predicting protein-ligand structures on challenging benchmarks and predicts small molecule binding affinities faster and cheaper than traditional physics-based methods.
• The engine can also accurately identify novel binding pockets on target proteins using only amino acid sequences, a capability previously requiring extensive experimental work.
• IsoDDE represents a leap forward in AI-driven drug discovery, enabling more precise and efficient design of new medicines by understanding biomolecular interactions with unprecedented accuracy.

• In July 2026, top-funded startups like Stripe ($33.7B), SpaceX ($11.8B), and ByteDance ($8.3B) demonstrate a trend of investor focus on companies controlling infrastructure, distribution, and workflow layers.
• Capital is being directed towards businesses that can establish hard-to-replace positions, particularly in AI compute, health, and developer tools.
• Founders are advised that a strong product alone is insufficient; a clear market entry, proven demand, and a product tied to recurring work are crucial for success.
• July 2026 funding activity indicates significant investment in AI compute, health software, fintech systems, and cloud-adjacent developer products.
• The key takeaway is to emulate the mechanism of successful startups by removing friction, owning a foundational layer, remaining lean, protecting assets, and building for customer retention, rather than mimicking their scale or style.

• 74% of U.S. employers utilize online tracking tools, and 75% monitor physical workplaces, indicating widespread employee surveillance in 2024-2025.
• 61% of U.S. companies employ AI for productivity or behavior analysis, and 67% collect biometric data.
• Workplace surveillance is linked to increased employee stress, with 45% reporting higher stress in high-surveillance environments compared to 28% in less monitored settings.
• A significant portion of employees express opposition to AI monitoring, with 61% against tracking movements and 56% against tracking desk presence.

• Employee monitoring is widespread, with 96% of companies using time-tracking software and 78% using other monitoring tools.
• A significant majority of employees (72%) disagree with employers' belief (68%) that monitoring improves work, with over half experiencing anxiety and privacy concerns.
• Employees are actively attempting to circumvent monitoring, with 49% faking online presence and 31% using anti-tracking tools.
• Future trends include AI predicting worker behavior, which 68% of employees oppose, and 54% would consider quitting if surveillance increased.

• San Francisco-based data and AI company Databricks has secured a $3 billion funding round at an $188 billion valuation, led by Coatue Management.
• This funding significantly increases Databricks' valuation, which has tripled in approximately 18 months.
• The capital will be used to expand AI products, fund acquisitions, and deepen AI research, focusing on enterprise AI infrastructure.
• Databricks is deferring an IPO, potentially until 2027 or later, citing market conditions and a lack of immediate need for public capital.

• US software and data analytics company Databricks announced yesterday (July 16) that it is raising a strategic funding round at an $188 billion valuation.
• The round is led by existing investor Coatue and is expected to include other new and existing investors, with a closing anticipated later this summer.
• The funding will accelerate Databricks' AI strategy by focusing on its Unity AI Gateway, Genie, and Lakebase offerings, and will also support future AI acquisitions and research.
• This valuation represents a significant increase from its February valuation of approximately $134 billion.

• Databricks announced strategic funding at a $188 billion valuation, with a term sheet signed and expected to close this summer, led by existing investor Coatue.
• The capital infusion will be used to accelerate Databricks' AI strategy, focusing on Unity AI Gateway, Genie, and Lakebase.
• These initiatives aim to address the "context gap" in enterprise AI by unifying data and AI, governing costs, and providing necessary business context for AI deployment.
• The funding will also support future AI acquisitions and deepen AI research.

• Databricks has secured a new funding round at an $188 billion valuation, led by Coatue Management.
• This valuation exceeds the $175 billion the company was reportedly in talks for previously.
• The funding round, reportedly around $3 billion, includes both new and existing investors and is expected to close this summer.
• Databricks offers a platform for data ingestion, analysis, and AI application development, with annual revenue exceeding $5.4 billion and growing over 65%.

• Databricks, a data analytics and AI company, has secured $3 billion in funding led by Coatue Management, reaching an $188 billion valuation.
• This valuation represents a 40% increase from its previous $134 billion valuation in February.
• The financing round, expected to close this summer, is driven by rapid revenue growth in Databricks' AI product suite, which achieved a $1.7 billion annualized revenue run rate in June.
• Funds will be used for AI initiatives, including new product development and strategic acquisitions, and the company has also expanded into cybersecurity with a recent acquisition.

• New research indicates that the software layer orchestrating AI agents, known as harnesses, significantly impacts token consumption and costs, more so than the AI model itself.
• Testing by AI consultancy Systima revealed that harness configurations, including system prompts and tool definitions, create substantial token overhead before user input.
• Academic research supports these findings, demonstrating that optimizing harness design can reduce token consumption, cost, and execution time by over 40% while maintaining output quality.
• CIOs are advised to focus on harness configuration and orchestration for cost control, rather than solely on model pricing, to gain visibility into the full economics of AI agent deployments.

• Deploying autonomous AI agents in production reveals that the true cost is significantly higher than advertised API token rates, with costs varying by use case and success rate.
• The critical metric for economic viability in 2026 is "Cost per Successful Task," which accounts for all expenses related to achieving a business outcome, not just API calls.
• A successful task is defined as a business KPI achievement that requires no human rework, such as a resolved support ticket or a qualified sales lead.
• The total cost per successful task includes run costs, human labor for oversight and correction, operational overhead, and amortized build costs.

• AI agent development costs in 2026 range from $8,000 for simple support chatbots to over $200,000 for complex multi-agent systems.
• Initial development represents only 25-35% of the total three-year cost, with ongoing expenses for tokens, maintenance, and evaluation making up the majority.
• Factors like autonomy, integrations, and compliance significantly influence pricing more than the underlying AI model.
• Renting an AI agent from a platform provider is often more cost-effective than building a custom solution if the workflow operates within a single system.

• AI agent performance exhibits predictable exponential decay patterns, allowing businesses to forecast capabilities and distinguish between successful and failed implementations.
• Benchmarks across various agent types, including computer interaction, remote browsing, and financial applications, reveal common limitations.
• Agents generally perform best in structured environments, with success rates declining as task complexity increases.
• Challenges persist in visual perception for screen interaction, adaptability to dynamic interfaces, and complex information retrieval, indicating that human oversight remains necessary for high-stakes tasks.

• The Federal Aviation Administration (FAA) has authorized Boeing to issue airworthiness certificates for all 737 MAX and 787 aircraft, effective next week.
• This decision follows a rigorous review of Boeing's production quality and safety data.
• The FAA will continue to oversee Boeing's certification process.
• The authorization aims to allow Boeing to increase production of these aircraft.

• The Federal Aviation Administration (FAA) will allow Boeing to issue airworthiness certificates for all 737 MAX and 787 airplanes starting next week.
• This decision follows months of FAA review and demonstrates confidence in Boeing's production quality under FAA oversight.
• The FAA had previously revoked Boeing's authority for these aircraft types due to production quality issues and fatal crashes.
• The FAA will continue to inspect and monitor Boeing's production system.

• Databricks announced on Thursday a new funding round valuing the company at $188 billion, led by Coatue.
• The company has undergone a significant valuation increase, transitioning from its big data roots to becoming a prominent AI provider.
• Databricks has recently launched several AI products, including Lakebase and Unity, and is known for its use of open-weight models.
• Internal benchmarking by Databricks indicated that open models and specific "harnesses" can manage AI costs effectively for complex coding tasks.

• Vertu's new $6,880 Alphafold foldable phone targets wealthy executives with a focus on an integrated AI agent, Hermes Agent, designed for task automation.
• The device features luxury materials like calfskin leather and titanium, but shares hardware similarities with the ZTE Nubia Fold.
• Hermes Agent demonstrates capabilities in analyzing local files and automating multi-step workflows, though it sometimes makes errors or requires clarification.
• The AI agent's performance is mixed when compared to Samsung's Gemini, with Hermes being more autonomous but less consistently accurate.
• Vertu emphasizes security features and the potential for enterprise integration, alongside the device's AI capabilities and luxury design.

• Australian founders have 48 hours left to apply for the Stripe x Startup Battlefield pitch competition, with applications closing July 20, 2026.
• Eight selected startups will pitch at Stripe Tour Sydney on August 19 for a chance to win prizes, including an automatic entry into TechCrunch Disrupt in San Francisco.
• The competition seeks promising, not necessarily polished, startups with a working MVP, and prior press or previous applications will not disqualify applicants.
• Successful applications will demonstrate a working product in real-time video, clearly identify competitors, and share the founders' compelling story.

• The macOS alert "Disk Not Ejected Properly" signifies that a storage device was unexpectedly disconnected, leading to its removal from the system's view.
• This occurs when the operating system's underlying hardware management (IOKit) loses track of the device, preventing proper unmounting and data integrity checks.
• Unlike a graceful unmount, this situation bypasses system notifications that allow applications to finish operations.
• The issue is often triggered by a sudden loss of connection, potentially related to the USB connection or device firmware.

• Union Pacific is applying white paint to its steel rails in high-heat areas to reduce surface temperature and prevent shifting.
• This technique, adapted from European rail practices, has shown a 20-degree drop in rail temperature and contributes to their ongoing safety efforts.
• The railroad achieved its best-ever full-year derailment incident rate in 2025, an improvement of 19% year over year.
• This initiative is part of a broader strategy that combines proven practices with continuous monitoring and innovation to strengthen the rail system.

• Kaiser Permanente nurses in California report that workplace surveillance, including the use of AI and strict call time monitoring, is negatively impacting their ability to provide patient care.
• Nurses state they face criticism or performance reviews for exceeding 15-minute call times, even in critical situations.
• AI systems are reportedly used to assess nurses' empathy and tone, raising concerns about patient privacy and professional judgment.
• The California Nurses Association is negotiating a new contract with Kaiser, with AI and workplace surveillance expected to be key issues.
• Kaiser Permanente maintains that its AI use prioritizes patient safety and that "average handle time" is not used to assess performance.

• The FAA announced Monday, July 20, 2026, that Boeing can resume issuing airworthiness certificates for all 737 MAX and 787 airplanes.
• This decision follows an eight-month review of production quality data and demonstrates the FAA's confidence in Boeing's ability to certify aircraft under its oversight.
• The FAA had previously restricted Boeing's ability to issue these certificates due to production quality issues.
• The FAA will continue to conduct inspections, audits, and monitoring of Boeing's production system.

• The Federal Aviation Administration (FAA) will allow Boeing to resume issuing airworthiness certificates for all newly built 737 MAX and 787 aircraft starting July 20, 2026.
• This decision returns a final production step to Boeing after eight months where the FAA and Boeing alternated this responsibility, finding comparable quality results.
• The FAA's decision does not recertify the 737 MAX or 787 as aircraft types and does not impact currently operating flights or existing tickets.
• Federal inspections, audits, and oversight of Boeing's production system will continue.

• The Federal Aviation Administration (FAA) has restored Boeing's authority to issue airworthiness certificates for its 737 Max and 787 Dreamliner jets, effective July 20.
• This authority was previously revoked due to manufacturing lapses and safety concerns, with regulators taking full control of 737 Max approvals in 2019 and 787 Dreamliner self-certification in 2022.
• The FAA's decision reflects confidence in Boeing's production processes under continued government oversight.
• Despite this restored authority, production limits on Boeing's 737 Max jets remain, though they have been gradually eased.

• Travel startup Fora has achieved a $1 billion valuation after securing $60 million in a funding round led by Forerunner and Tactile Ventures.
• Launched in 2020, Fora provides independent travel advisors with a platform to manage bookings and client relationships, catering to increased traveler demand for personalized service.
• The platform now serves over 15,000 advisors and has facilitated more than $3 billion in bookings, with plans to incorporate AI and expand its advisor base.

• Fora, a platform for travel entrepreneurs, announced it has raised $60 million in a Series D funding round at a $1 billion valuation.
• The company has achieved over $3 billion in bookings since its 2021 founding and is rolling out an embedded AI assistant named Via to support travel advisors.
• The new funding will be used to enhance AI capabilities, expand into new markets, and grow its presence in various travel sectors.
• Fora's model attracts new advisors, with 97% of its over 15,000 active advisors being new to the profession.

• Fora, a travel advisor platform, has raised $60 million in Series D funding, achieving a $1 billion post-money valuation.
• The New York-based company will use the funds to accelerate AI development, specifically its AI assistant Via, and expand into new travel categories.
• Fora's network of advisors has generated over $3 billion in travel bookings since its 2021 launch, with recent growth accelerating significantly.
• The platform aims to enhance the role of travel advisors by automating operational tasks, allowing them to focus on personalized client service.
• Fora has over 15,000 active advisors, 97% of whom are new to the profession, and plans to expand its offerings in cruise, flights, and enterprise.

• AI-powered travel agency Fora has achieved unicorn status, raising $60 million in a Series D funding round that values the company at $1 billion.
• The platform empowers travel entrepreneurs and has seen rapid growth since its 2021 founding, with advisors booking over $3 billion in travel.
• Fora is developing an AI assistant, Via, to enhance the capabilities of travel advisors by automating tasks like research and itinerary creation.
• The funding will be used to further develop AI features, expand into new markets, and grow in various travel sectors.

• Generative AI companies secured $271.01 billion across 84 funding rounds between Q2 2025 and Q2 2026.
• Foundation-model companies received 95.0% of the total capital despite accounting for only 13 deals.
• OpenAI, Anthropic, and xAI were responsible for the largest funding rounds, with OpenAI's $122 billion financing being the largest single round.
• Enterprise generative AI applications had the most deals (33) but received only 2.0% of the total funding.

• US power demand from AI data centers could grow thirtyfold to 123 gigawatts by 2035, creating significant strain on existing infrastructure.
• Hyperscalers are building increasingly large data centers, with some future projects potentially consuming 5 GW, equivalent to five million homes.
• AI data centers create concentrated, 24/7 power demands that challenge grid operations, leading to issues like harmonic distortions and generation shutdowns.
• A Deloitte survey found grid stress to be the leading challenge for data center infrastructure development, with 79% of respondents expecting AI to increase power demand.
• Long grid connection wait times, supply chain disruptions, and security concerns further complicate the expansion of AI infrastructure.

• U.S. hyperscalers are projected to spend up to $1.12 trillion on AI-related capital expenditures by 2027, with a significant portion directed towards infrastructure like cooling systems and power plants.
• Industrial companies Vertiv and Argan are positioned to benefit from this AI spending boom.
• Vertiv provides power and cooling equipment for data centers and has a substantial backlog of orders.
• Argan constructs power generation facilities and also supplies components for cooling systems.

• Major tech companies, including Microsoft, Amazon, Google, and Meta, are entering into nuclear power deals to meet the immense and constant electricity demands of AI data centers.
• AI's power consumption is rapidly increasing, with global data center electricity use projected to nearly double by 2030, and U.S. data centers potentially consuming 12% of national electricity by 2028.
• Nuclear power is favored for its steady, round-the-clock output and low-carbon emissions, addressing limitations of grid capacity, and the intermittency of renewables.
• Deals include restarting existing plants for faster power delivery and funding new Small Modular Reactors (SMRs), though new SMRs are largely expected in the late 2020s and 2030s.
• The rising cost and scarcity of power for AI will likely translate to higher AI service prices and potential capacity constraints for users.

• AI model-routing marketplace OpenRouter is reportedly fielding takeover interest from larger tech companies.
• The company recently closed a $113 million Series B funding round in May 2026, valuing it at approximately $1.3 billion.
• OpenRouter serves around 8 million users, processes trillions of tokens monthly, and had an estimated $50 million in annualized revenue by early 2026.
• The platform offers access to over 400 AI models through a single API, allowing users to switch between them.

• Xgimi has introduced two new portable laser projector models to its Elfin line.
• Both the Elfin Flip Laser and Elfin Flip 4K feature a triple laser light source, increasing brightness to 1,600 ISO lumens from the original 500 ISO lumens.
• The Elfin Flip Laser is priced at $799 and supports 1080P resolution, while the Elfin Flip 4K costs $999 and includes 4K resolution.

• AI-powered travel agency Fora has achieved unicorn status, raising $60 million in a Series D funding round led by Forerunner and Tactile Ventures, valuing the company at $1 billion.
• Founded in 2021, Fora operates a two-part platform enabling individuals to become travel agents and allowing users to find advisors for trip planning.
• The new funding will support the expansion of Fora's AI assistant, Via, to help agents with administrative tasks, aiming to enhance productivity rather than replace human advisors.
• Agents on the Fora platform have booked over $3 billion in travel since its launch, with the company planning to use the funds for hiring and growth in new travel categories.

• The Federal Aviation Administration (FAA) has authorized Boeing to resume issuing airworthiness certificates for its 737 Max and 787 Dreamliner aircraft.
• This authority was previously removed following fatal 737 Max crashes in 2018 and 2019.
• The FAA's decision comes after an eight-month period where both Boeing and the FAA issued certificates, showing comparable production quality findings.
• The agency stated it can safely return this responsibility to Boeing based on these results.

• The Department of Energy's Office of Nuclear Energy has identified six advantages and five challenges of using nuclear energy to power data centers.
• Advantages include 24/7 reliability, long operational periods without refueling, flexible reactor designs, stable costs, compact footprint, and potential use of existing or retired nuclear plants.
• Challenges involve the lengthy time and high initial cost for new reactor construction, potential metering issues for co-located facilities, an underdeveloped domestic nuclear fuel supply chain, and the need for spent fuel storage solutions.
• The DOE is supporting efforts to scale up nuclear fuel production and exploring advanced technologies to manage spent fuel.

• Tech giants are increasingly using nuclear energy to power AI data centers due to the high, constant electricity demands of AI processors and cooling systems.
• Companies are securing nuclear power through co-location with existing plants, long-term power purchase agreements, and direct investment in startups developing Small Modular Reactors (SMRs).
• Nuclear energy provides reliable, zero-carbon baseload power that helps tech companies meet their carbon-neutrality goals while overcoming limitations of renewable sources and strained power grids.
• The development and deployment of SMRs are seen as a key future strategy, offering compact, modular, and advanced safety features for dedicated AI data center power.

• X-energy is partnering with Amazon to deploy 5 gigawatts of its Xe-100 advanced nuclear reactors by 2039 to power artificial intelligence data centers.
• Exponential growth in AI is creating unprecedented demand for reliable, continuous energy, straining traditional power sources and leading to significant projected increases in U.S. electricity demand by 2029.
• Data centers supporting AI workloads are identified as the primary driver of this increased demand, with single AI queries consuming substantially more electricity than standard web searches.
• X-energy's Xe-100 reactors offer clean, reliable power with a small land footprint and high output stability, making them suitable for AI infrastructure needs.

• In late 2024, AI companies began making deals with the nuclear industry to secure power for data centers.
• Microsoft is investing in restarting the Three Mile Island plant, while Google and Amazon are investing in small modular reactors (SMRs) for future data centers, though SMRs are currently unproven in the Western world.
• Significant challenges for nuclear power include high costs, construction delays, and the unproven viability of SMRs, while AI's energy demand may not escalate as predicted due to technological advancements.
• Governments are advised to invest cautiously in nuclear power, prioritizing proven energy sources like wind, solar, and geothermal in the interim.

• US nuclear capacity is projected to triple by 2050, driven by the energy demands of artificial intelligence and data centers.
• Nuclear power offers a consistent, carbon-free energy supply essential for AI operations, unlike intermittent renewable sources.
• Major tech companies like Microsoft and Google are investing in nuclear energy to power their data centers, with Microsoft purchasing output from a revitalized nuclear facility and Google planning to acquire energy from small modular reactors.
• While nuclear power presents advantages for AI infrastructure, challenges remain, including high construction costs, regulatory hurdles, and public concerns.

• Global AI infrastructure spending reached $89.9 billion in Q4 2025, a 62% year-over-year increase, concluding a record year.
• Full-year 2025 spending totaled $318 billion, more than doubling 2024's $153 billion, driven by U.S. hyperscaler investment and server adoption.
• The United States accounted for 77% of global spending, though China experienced an 8.1% decline due to export controls.
• Emerging regions, particularly the Middle East & Africa, saw significant growth, exceeding 500%.
• Global AI infrastructure market is projected to surpass $1 trillion by 2029.

• AI companies are outbidding carmakers for DRAM memory chips, leading to a new global chip crunch.
• Spot prices for DRAM have surged approximately 450% between September 2025 and January 2026.
• Carmakers like Ford and GM are increasing procurement forecasts by hundreds of millions of dollars due to rising costs.
• While some automakers report stable supply chains and no production adjustments yet, the demand from AI data centers and limited capacity expansion suggest pricing pressure will persist.

• Memory producers are prioritizing High Bandwidth Memory (HBM) for AI servers, leading to reduced availability and higher prices for general DRAM in 2026.
• This shift is driven by higher profit margins and demand for advanced computing, impacting data centers, automotive, industrial electronics, and consumer devices.
• The shortage is prompting changes in chip design, with a move towards chiplets and advanced packaging, and has led semiconductor companies to expand capacity and diversify manufacturing locations.
• While supply is expected to gradually improve as new production comes online, high-end memory demand is likely to remain strong throughout 2026.

• Amazon Web Services (AWS) will invest up to $50 billion to build the first purpose-built AI and supercomputing infrastructure for U.S. government agencies, with construction beginning in 2026.
• This expansion will add approximately 1.3 gigawatts of compute capacity across AWS's Top Secret, Secret, and GovCloud regions.
• The investment aims to provide federal agencies with expanded access to AWS AI services and secure infrastructure to advance American AI leadership and accelerate mission-critical tasks.

• NVIDIA and Intel have announced a collaboration to develop custom data center and personal computing products.
• The partnership will integrate NVIDIA's AI and accelerated computing technologies with Intel's CPUs and x86 ecosystem.
• Intel will build custom x86 CPUs for NVIDIA's AI infrastructure and x86 system-on-chips (SOCs) integrating NVIDIA RTX GPU chiplets for PCs.
• NVIDIA will invest $5 billion in Intel's common stock as part of the agreement.

• Meta is reportedly in early discussions to rent up to $10 billion worth of AI compute power to its competitor, Anthropic, over two years.
• This potential deal, first proposed by Anthropic in June, is being explored as Meta has built AI capacity exceeding its current model development needs.
• The arrangement would be overseen by a new Meta Compute unit and could involve offering raw GPU capacity or a hosted access layer for models.
• This move follows a similar agreement where Anthropic leased a data center from SpaceX.

• Chip equipment manufacturer ASML is awarding approximately 44,500 employees worldwide €20,000 in shares, with a restriction on selling them before January 1, 2030.
• This announcement follows ASML's strong Q2 2026 financial results, including a €2.9 billion profit on €9.3 billion in sales, and a second raised full-year forecast driven by AI demand.
• The share grant is notable as it contrasts with earlier job cuts and a union agreement preventing forced redundancies until May 2027.

• Inference cloud startup General Compute secured a $400 million loan from Upper90, using SambaNova SN50 chips as collateral.
• This marks a potential shift in AI infrastructure financing, with debt backed by non-Nvidia inference-specific silicon.
• The loan facility begins at $100 million and scales with customer demand, with the air-cooled SN50 infrastructure deployable in weeks.
• This deal could provide a template for other inference chipmakers to secure financing without diluting equity.

• Roughly 300 Netflix titles have utilized generative AI, primarily in post-production, according to the company's second-quarter earnings report released Thursday.
• Examples include the docuseries The American Experiment, which featured 17 minutes of "AI-enhanced footage" created twice as fast and at half the cost.
• Netflix is investing in AI for faster, cheaper, and higher-quality content production, acquiring an AI startup and establishing an animation studio.
• The company also noted its ad revenue is on track to double and addressed viewer engagement, emphasizing quality and variety alongside watch time.

• Nuclear startup Valar Atomics is reportedly in discussions to raise new funding at a $6 billion valuation.
• Sequoia Capital is expected to lead the funding round, which aims to raise $1 billion in equity.
• Valar Atomics develops small modular nuclear reactors (SMRs) and recently demonstrated its reactor powering an Nvidia AI chip, announcing a partnership to explore nuclear energy for AI data centers.
• The company's technology utilizes a helium-cooled, high-temperature gas reactor design, with plans to build hundreds of SMRs for data centers.

• Lisp is a family of programming languages with over 20 dialects, differing in operators, semantics, and capabilities, but sharing a fundamental syntax.
• Common Lisp (CL) is a mature, standardized dialect (1994 ANSI specification) known for its comprehensive features, native code compilation via implementations like SBCL, and a powerful condition/restart debugging system.
• CL supports multiple programming paradigms, including object-oriented programming with its advanced CLOS system, and offers stability with backward compatibility.
• Clojure, designed for the JVM, offers functional programming with immutable data structures, built-in concurrency support, and efficient persistent data structures.
• Clojure's host platform integration (JVM and JavaScript via ClojureScript) allows for full-stack development, though debugging can be complicated by platform-specific details.

• Running SQLite for a Django site revealed the need for database maintenance, particularly the ANALYZE command which drastically improved slow full-text search queries.
• Database cleanup operations involving deleting many rows can cause timeouts and worker crashes if they exceed the database's write timeout.
• The author explored two backup methods: restic with VACUUM and gzip, and litestream for incremental backups.
• It's possible to use multiple SQLite database files to separate tables, a technique previously used successfully in another project.

• U.S. companies signed approximately $60 billion in deals with Iraq on Friday, including agreements aimed at developing alternative oil shipping routes.
• The agreements, signed at the U.S. Chamber of Commerce, also cover healthcare, communications, and infrastructure sectors.
• Among the deals, Chevron signed three agreements with the Iraqi government, two focused on increasing oil production and one for a new export pipeline.
• The State Department also welcomed an agreement between Iraq and Syria to rehabilitate their crude oil pipeline.
• The potential timeline for these alternative routes to become viable is unclear, with estimates suggesting pipeline construction through multiple countries could take at least two and a half years.

• Artificial intelligence company Anthropic is reportedly in negotiations with Meta to lease computing power in a potential two-year deal valued up to $10 billion.
• The discussions, which began in June, are still in early stages, and an agreement is not guaranteed.
• This potential arrangement could mark a new revenue stream for Meta, which has indicated interest in expanding into cloud computing to monetize its AI investments.
• Anthropic is seeking to increase its computing capacity, a challenge due to shortages of Nvidia chips.

• SpaceX stock has fallen below its Initial Public Offering (IPO) price of $135 for the first time.
• The company has lost approximately $1 trillion in market value since its peak on June 16.
• A recent aborted Starship rocket launch in South Texas contributed to the stock's decline.
• Looming lockup expirations for employees and early investors are also impacting the stock price.

• Cleveland Fed President Beth Hammack stated Friday that interest rates may need to rise to combat persistent inflation, a view echoed by other policymakers expressing concern about rising fuel prices and AI data center build-outs.
• Hammack noted that businesses are now suggesting the Fed take action on inflation and consumers are expressing despair over affordability.
• Dallas Fed President Lorie Logan and Fed Vice Chair Philip Jefferson have also indicated that higher interest rates could be appropriate if inflation does not cool down.
• Traders currently estimate a 15% chance of a July rate hike, with that probability increasing to 65% by the September meeting.

• The Philadelphia SE Semiconductor Index has fallen approximately 9% this week, marking its largest weekly decline in over a year and a nearly 20% drop from its late June high.
• This pullback in chip stocks, which have been key drivers of the AI rally, is raising concerns about the sustainability of the gains and investor overextension.
• Factors contributing to the selloff include profit-taking, scrutiny of AI capital expenditure sustainability, and reports of delays in AI model releases from companies like Google.
• Other market indicators, including momentum stocks and global indices like South Korea's KOSPI and Japan's Nikkei, have also experienced recent weakness despite year-to-date gains.

• Gabriel Perez, President Trump's teleprompter operator, is under investigation for allegedly using confidential information to bet on at least a dozen presidential speeches.
• Perez reportedly profited over $100,000 on the prediction market platform Kalshi, which flagged suspicious activity to the Commodity Futures Trading Commission (CFTC).
• Perez has been placed on unpaid administrative leave, and President Trump called the situation a "disgrace."
• The White House stated it has strict ethics guidelines and is unaware of any other staffers misusing confidential information for betting.

• The Commodity Futures Trading Commission (CFTC), a Trump administration regulator, has ordered the prediction market Kalshi to defy a Michigan court order to cancel pending sports trades.
• This action, utilizing emergency powers not seen since the Cold War, represents an escalation in federal efforts to prevent states from regulating prediction markets.
• Kalshi stated it had already complied with the Michigan judge's order, potentially making the immediate legal standoff moot, but the CFTC's maneuver is seen as its most aggressive step to date.
• The CFTC's intervention aims to prevent a state court ruling from setting a precedent in ongoing federal lawsuits against states attempting to regulate prediction markets.

• Two Utah residents, Gabi Finlayson and Josh Kanter, have filed a lawsuit against Kevin O'Leary and Fox News.
• The lawsuit alleges O'Leary defamed them on Fox News by claiming their opposition to his Utah data center was funded by the Chinese government.
• Finlayson and Kanter claim these allegations caused reputational harm, economic losses, emotional distress, and safety threats.
• O'Leary previously issued a social media statement clarifying he had no evidence to support his claims of Chinese funding against the plaintiffs.

• CVS Health CEO David Joyner stated that private insurers will not cover GLP-1 drugs for weight loss until their prices decrease.
• Joyner explained that current prices for GLP-1s do not demonstrate sufficient downstream health savings to justify coverage by insurers.
• Private insurance generally declines to cover these drugs for weight loss unless they are used to treat other medical conditions, such as Type 2 diabetes.
• The high cost of GLP-1s, which can range from hundreds to thousands of dollars monthly without insurance, is a significant factor in this coverage stance.

• Oracle reported Q4 fiscal 2026 revenue of $19.2 billion, with cloud infrastructure revenue growing 93% and cloud applications revenue increasing 10%.
• The company's growth is attributed to demand for AI infrastructure and applications, with executives highlighting AI roadmap developments across their technology stack.
• Oracle is integrating over 1,000 AI agents into its application suites and emphasizing the use of enterprise data for AI use cases and enhanced database security.
• Oracle Cloud Infrastructure (OCI) is experiencing strong demand for AI infrastructure, evidenced by significant contract signings and high GPU utilization.
• The company is exploring pricing models tied to agentic AI usage and business outcomes to connect AI spending with specific application benefits.

• Oracle reported Q4 FY 2026 total revenue of $19.18 billion, up 21% year-over-year, driven by a 93% YoY increase in cloud infrastructure revenue to $5.79 billion.
• AI infrastructure demand significantly boosted Oracle's backlog, with $67 billion in AI infrastructure contracts signed in the quarter.
• The company is packaging agentic AI capabilities for simpler commercial structures and reported strong multicloud database growth of 404% YoY.
• Oracle provided guidance for Q1 FY 2027 total revenue growth of 27%-29% and raised FY 2027 non-GAAP diluted earnings per share guidance.

• A client is experiencing technical difficulties with a website.
• The website states that JavaScript is disabled in the user's browser.
• Users are instructed to enable JavaScript, check their connection, disable ad blockers, or try a different browser to resolve the issue.

• Fidelity Investments has reached a settlement in a class action lawsuit, In re: Fidelity Investments Data Breach Litigation, regarding a data security incident that occurred between August 17 and August 19, 2024.
• A third party gained unauthorized access to information on Fidelity's computer network during this period.
• Fidelity denies wrongdoing, and the settlement aims to resolve the litigation without a court decision on the merits.
• Individuals identified in Fidelity's records may be eligible for benefits as "Settlement Class Members."
• Key deadlines include objections and opt-outs by June 26, 2026, and claims by July 27, 2026, with a final approval hearing scheduled for July 9, 2026.

• Fidelity Investments has agreed to a $2.5 million class action settlement for a data breach that occurred in August 2024.
• Individuals whose financial account and routing numbers were compromised in the breach, or who received a notice from Fidelity regarding the incident, may be eligible for compensation.
• Eligible class members can claim up to $5,000 for documented losses, a pro rata cash payment (estimated at $100), and two years of identity theft protection.
• California residents may be eligible for an additional payment.
• The deadline to file a claim is July 27, 2026.

• Fidelity Investments has agreed to a $2.5 million settlement to resolve a class-action lawsuit stemming from a data breach that occurred between August 17-19, 2024.
• The breach exposed the names and personal identifiers, including Social Security numbers, of over 77,000 customers, and potentially impacted an additional 86,000 customers with compromised financial account information.
• Impacted customers can file a claim on the settlement website by July 27 to potentially receive up to $5,000.
• The settlement requires court approval, with a final hearing scheduled for July 9.

• A nationwide cyclosporiasis outbreak has been traced to iceberg lettuce from Mexico used by Taco Bell.
• The Food and Drug Administration identified a single supplier, reportedly Taylor Farms, for the contaminated lettuce.
• The outbreak has sickened at least 1,644 people across 34 states, with many hospitalizations.
• Taco Bell has voluntarily removed the implicated lettuce and is replacing it nationwide.

• The US Food and Drug Administration has identified shredded iceberg lettuce supplied to Taco Bell as a source of cyclospora, a parasite causing an ongoing outbreak of watery diarrhea.
• The affected lettuce was served at Taco Bell locations in Indiana, Kentucky, Michigan, Ohio, and West Virginia, linked to 1,644 confirmed cases.
• The FDA investigation pointed to a single supplier of iceberg lettuce from Mexico, with a person familiar with the matter identifying the supplier as Taylor Farms.
• Taco Bell has voluntarily removed the potentially impacted lettuce from its supply chain nationwide and is replacing it.
• Cyclospora infections have spread to 34 states, with over 1,600 cases confirmed nationally.

• Lettuce supplier Taylor Farms is preparing to recall ingredients linked to a parasite outbreak sickening thousands in Michigan and nearby states.
• The recall would be the clearest indication of the source of a widening outbreak of severe diarrhea tied to shredded lettuce.
• US health authorities identified shredded lettuce from a single supplier as the cause of the illnesses.
• The Centers for Disease Control and Prevention has confirmed illnesses in 34 states, with Michigan reporting over 5,000 cases.

• Apple is reportedly in early settlement talks with the U.S. Department of Justice (DOJ) regarding an antitrust lawsuit.
• The discussions are aimed at resolving the legal dispute without a full trial.
• Bloomberg News reported on these ongoing negotiations.

• Apple and the US Department of Justice (DOJ) are reportedly in settlement talks regarding the DOJ's antitrust lawsuit, filed in 2024, which accuses Apple of a "smartphone monopoly."
• Apple has reportedly made multiple offers this year to resolve the case, though a settlement is not guaranteed.
• The lawsuit focuses on aspects of the iPhone experience including super apps, cloud streaming games, third-party messaging, smartwatches, and digital wallets.
• Apple has made recent changes, such as supporting RCS and broadening NFC chip access, in an attempt to address some of the DOJ's concerns.

• Apple and the US Justice Department are in early settlement discussions regarding a 2024 antitrust lawsuit.
• The lawsuit alleges Apple violated antitrust laws by harming competitors, developers, and consumers.
• Apple has made several offers to resolve the case, but no agreement is guaranteed, and no trial date is set.
• The Justice Department's allegations include blocking super apps, hindering messaging and cloud streaming, and limiting smartwatch competition.
• Apple has since made some adjustments, such as opening its Messages app to RCS and allowing third-party access to its payment chip.

• Since 2023, artificial intelligence has become a significant driver of US economic growth, contributing to approximately half of the increase in real GDP.
• Productivity growth has accelerated, though these gains are primarily concentrated in technology-related industries.
• Despite productivity increases, broad-based job losses due to AI have not yet been observed; employment in AI-exposed occupations continues to grow.
• Tech sector employment has declined, but this is attributed to post-pandemic adjustments rather than widespread AI displacement.

• Emoji usage has evolved from literal interpretations to conveying nuanced emotions and subtext in digital communication.
• Expressions like "face with tears of joy" (😂) are declining in popularity, replaced by more hyperbolic and intense emotions such as "loudly crying face" (😭) and "rolling on the floor laughing" (🤣).
• Emoji design is shifting to a 3D style, prioritizing expressiveness and character over photorealism, with user studies informing design choices.
• New AI tools are being used to ensure emoji with darker skin tones are visible in dark mode by improving contrast ratios.
• The new Noto 3D emoji are open-source, allowing the community to use and remix them for various applications.

• Frank Lloyd Wright's Oak Park Home and Studio, where he lived and worked from 1889 to 1909, is a significant architectural landmark in Illinois.
• The home, initially built by Wright at age 22, underwent several expansions, including the addition of a studio wing in 1898, to accommodate his growing family and burgeoning independent architectural practice.
• Many of Wright's early Prairie-style designs were conceived in this Oak Park studio before he moved on to establish residences in Wisconsin and Arizona.
• After falling into neglect, the property was purchased and restored by the Frank Lloyd Wright Home and Studio Foundation and the National Trust for Historic Preservation, reopening for public tours in 1974.
• The home, originally a shingle-style structure, showcases early examples of Wright's design innovations, such as barrel-vaulted ceilings and the concept of "compression and release."

• Apple and the U.S. Justice Department are in early settlement talks regarding the 2024 antitrust lawsuit that accuses the company of violating antitrust laws.
• Discussions are active, but there is no guarantee of an agreement.
• The Justice Department's lawsuit, filed alongside 19 states and the District of Columbia, alleges Apple harmed competitors and consumers through practices like blocking super apps and discouraging rival messaging and digital wallet solutions.
• Apple has reportedly made offers to resolve the case and has already implemented some changes that address the DOJ's concerns.

• Lettuce supplier Taylor Farms is preparing to recall ingredients linked to a parasite outbreak that has sickened thousands in Michigan and nearby states.
• The recall by the major produce supplier is the clearest sign yet of the potential source of the outbreak, which US health authorities have tied to shredded lettuce.
• Taco Bell had previously removed some lettuce from Midwest restaurants due to the outbreak.

• Oracle's stock has fallen 35% year-to-date, with Guggenheim maintaining a $400 price target, suggesting potential for the stock to more than triple in 12 months.
• The company's stock decline follows S&P cutting its credit rating to BBB- and concerns over significant AI infrastructure spending impacting free cash flow.
• Oracle recently reported negative free cash flow for FY2026 and has a substantial contracted backlog, which Guggenheim believes the market is overlooking.

• Sweetgreen Inc. shares rose significantly on Friday, halting a four-day decline.
• This surge follows a US regulatory announcement tracing a parasite outbreak to shredded iceberg lettuce at Taco Bell restaurants.
• Previously, Sweetgreen's stock had dropped due to concerns that the outbreak would reduce demand for fresh produce.
• The CDC identified the outbreak's source, and a lettuce supplier is preparing a recall.

• U.S. companies have received approximately $71 billion in tariff refunds after a Supreme Court ruling invalidated certain tariffs.
• The refunds, with $49.2 billion issued in June, are being used by companies like PepsiCo and McCormick & Company to offset rising commodity costs.
• These increased costs are attributed to inflation driven by factors including the Iran war and its impact on gas prices, alongside supply chain issues.
• Analysts express concern that renewed geopolitical tensions in the Middle East could lead to further price increases.

• A $2.5 million settlement for a data breach affecting Fidelity Investments customers has a claim deadline of July 27, 2026.
• The breach occurred between August 17 and August 19, 2024, compromising personal data for 77,099 individuals.
• Eligible customers can receive up to $5,000 for documented losses, plus two years of identity theft protection and credit monitoring.
• Claims can be submitted online at FidelityDataSettlement.com/form/claim or by mail before the deadline.

• Taiwan Semiconductor Manufacturing Co. (TSMC) has raised its capital expenditure and revenue targets, citing strong demand for artificial intelligence.
• ASML Holding plans to expand its capacity by up to 70% by 2028 due to firm demand.
• Company reports highlight trends in the technology sector, including Wipro's concerns about long-term growth and Ericsson's dominant position in mobile networking despite profitability delays.
• SK Hynix's ADRs have begun trading under a new ticker, with no anticipated shift in long-term dynamics.
• Other companies like Grab and CrowdStrike are featured with reports on their respective market positions and strategies.

• A new private jet fractional ownership company, Bond, has launched in Fort Lauderdale, Florida, aiming to cater to the growing number of ultrawealthy individuals.
• The surge in AI-driven wealth is contributing to a scarcity of private jets.
• Bond's founder previously sold his private jet charter business for nearly $1 billion in 2022.

• SpaceX (SPCX) shares fell over 3% in extended trading on Thursday after its Starship mega rocket launch was aborted.
• The launch attempt was called off, though CEO Elon Musk indicated a retry is possible in a few days.
• The stock's decline pushes it further below its initial public offering (IPO) price of $135.

• SpaceX's stock has fallen 25% since its June IPO, following concerns about the company's high valuation compared to its substantial losses in launch and AI divisions.
• The company's profitable Starlink internet service is reportedly subsidizing its money-losing AI and space launch operations.
• Debt markets have repriced SpaceX's risk, with its investment-grade debt trading like junk, and a significant portion of insider shares are set to unlock, potentially adding further downward pressure.
• Analyst opinions are divided, with some predicting further declines due to unachieved revenue goals, while others forecast substantial gains based on the company's expansive market vision.

• SpaceX stock has dropped significantly since its initial public offering (IPO), losing approximately $1 trillion in market value from its June 16 peak.
• The company's shares have fallen below their IPO price of $135, closing at $131.11 on Thursday and trading below $125 by midday Friday.
• The stock's decline follows the cancellation of a planned Starship launch due to engine issues, with a rescheduled attempt expected early next week.
• Looming lockup expirations, allowing employees and early investors to sell a substantial number of shares, are also contributing to the stock's downward pressure.

• SpaceX shares fell 3% on Thursday to an all-time low of $131, trading below its $135 IPO price for the first time.
• The company scrubbed its 13th Starship test flight due to a booster issue, with no new launch date announced.
• This marks the fifth consecutive day of stock declines, as investors weigh Starship's development against execution timelines.

• Gina Rinehart's company, Hancock Prospecting, may have seen an estimated A$700 million paper gain from its SpaceX IPO investment wiped out as the stock price dropped below its debut price.
• SpaceX shares closed at $131.11 on Thursday, down from their IPO price of $135, and the company's market value has fallen to $1.72 trillion from over $2.6 trillion.
• Hancock Prospecting has not commented on the current value of its SpaceX stake.
• Approximately 28,000 Australians applied to buy SpaceX shares, with some investors having already sold to secure profits while others anticipate long-term returns.

• A Q2 2026 analysis by J.P. Morgan identifies trends in venture capital, noting that high-profile mega-rounds are shaping the market.
• Venture capital strategies are diverging, with capital increasingly concentrating in AI startups.
• The 2026 IPO market is projected to be historic, alongside increased M&A activity.
• Persistent higher interest rates and policy uncertainties continue to elevate the cost of capital.

• The Bay Area in the United States is the leading region for venture capital, with $390.3 billion raised in the last 12 months ending Q2 2026.
• The United States remains the largest source of venture capital, followed by Europe and Asia, with China, India, Singapore, and Israel anchoring Asian investment.
• Globally, OpenAI, Anthropic, and xAI are among the top funded companies, based on total funding.
• In terms of growth, Duolingo, MercadoLibre, and BYD are ranked highest according to Dealroom's growth indicator.

• China's AI agent regulations, requiring a three-tier decision authorization framework and mandatory filings, became enforceable on July 15, 2026.
• New actions are required for organizations, including confirming compliance with China's agent rules, blocking the xAI Grok Build CLI due to data transmission risks, implementing two-person verification for AI-generated legal content, and adding input validation to financial AI agents.
• Organizations should monitor Illinois's third-party safety audit mandate, DHS-CISA security rules for critical infrastructure AI, OpenAI's pre-release evaluation proposals, and UN Global Dialogue on AI Governance standards.
• Trends indicate agentic AI risks are accelerating, outpacing governance, and regulatory compliance is becoming increasingly complex with simultaneous enforcement across jurisdictions.

• In the second quarter of 2026, U.S. lawmakers introduced multiple AI-related bills focusing on chatbots, digital replicas, and nonconsensual imagery, while the executive branch issued an executive order on AI security and the Department of Justice intervened in a state AI law challenge.
• States enacted numerous AI bills, with new laws in Illinois and Connecticut addressing frontier model oversight, and other states passing legislation on automated decision-making in employment, chatbot safety, and AI use in healthcare.
• Developments in connected and automated vehicles included updated California regulations for AV testing, House approval of a bill for autonomous trucking, and NHTSA's proposal to amend brake system standards.
• The Internet of Things saw the introduction of the GUARD Act to review foreign robots for national security risks and the naming of ioXt Alliance as the lead administrator for the FCC's Cyber Trust Mark program.

• Over 300 artificial intelligence products will debut at the World AI Conference (WAIC) in Shanghai starting Friday, with a significant focus on practical applications.
• The conference will feature over 1,100 companies and 3,000 exhibits, demonstrating AI's integration into manufacturing, healthcare, education, and daily life.
• During the event, 29 countries signed an agreement to establish the World Artificial Intelligence Cooperation Organization (WAICO) headquartered in Shanghai, aiming to promote beneficial, safe, and fair AI development.
• Showcased debuts include advanced humanoid robot hands, heavy-duty embodied robots, AI-powered retail solutions, and new AI models and computing infrastructure.

• The first week of July 2026 saw the release of new AI models from SpaceXAI, OpenAI, and Meta, triggering a significant price war that drastically reduced inference costs.
• DeepSeek announced plans to develop custom inference silicon, aiming to reduce reliance on Nvidia and Huawei.
• Microsoft implemented 4,800 job cuts amid financial struggles in its Xbox division, while its outgoing CEO joined a Federal Reserve task force on AI productivity.
• Starbucks began developing internal AI software, signaling a potential disruption to the traditional enterprise SaaS model.

• Images and specifications for Samsung's upcoming Galaxy Z Fold 8, Galaxy Z Fold 8 Ultra, and Galaxy Z Flip 8 have been leaked by Evan Blass, ahead of an expected July 22nd launch.
• The leaked images depict a redesigned Galaxy Z Fold 8 with a shorter, wider display, powered by a Snapdragon 8 Elite For Galaxy processor and featuring dual 50MP rear cameras.
• The Galaxy Z Fold 8 is expected to offer up to 26 hours of video playback, while the Z Fold 8 Ultra may include a 200MP wide rear camera and a 5,000mAh battery.
• The Z Flip 8 leak suggests it will retain the current Z Flip 7 camera setup with a 4,300mAh battery.

• Bunkerhill Health has raised $55 million in a Series B funding round to expand its agentic AI platform, Carebricks.
• The platform allows health systems to build custom AI agents for clinical and administrative tasks, aiming to bridge the gap between AI models and live patient data at scale.
• Early adopters like Cleveland Clinic, UTMB, and Intermountain Health are using Carebricks, with UTMB deploying over 20 agents that have reportedly improved patient care and operational efficiency.
• The funding will be used to broaden Carebricks' applications and enhance its governance, monitoring, and safety features.

• A TechCrunch Disrupt 2026 session in San Francisco will focus on how founders can secure pre-seed funding even without a product.
• The panel, titled "Winning Pre-Seed Without a Product," aims to address challenges faced by founders due to increased seed funding for AI startups.
• Speakers include Sandhya Venkatachalam (Axiom Partners), Puneet Agarwal (True Ventures), and Austin Clements (Slauson & Co.), who will share expertise on investor expectations and fundraising strategies.
• The session is part of Disrupt 2026, taking place October 13-15 at Moscone West.

• An AI auditor identified a critical soundness bug in OpenVM's guest library, openvm-pairing, allowing malicious provers to forge pairing equalities.
• The bug, assigned CVE-2026-46669, was fixed in OpenVM 1.6.0, and partners have since upgraded.
• This issue was found by the AI auditor zkkao after a more specialized context engineering approach was used, as simpler LLM setups failed to identify exploitable bugs in the complex zkVM codebase.
• Forging pairing checks can compromise various cryptographic applications, including SNARK verifiers, BLS signatures, and Ethereum ecPairing precompiles executed within OpenVM.

• Tannakian reconstruction is a method to recover the structure of a source category from its image in a target category with sufficient "resolution."
• This process is analogous to superimposing multiple long-exposure photographs to pinpoint a specific light source obscured by other lights.
• The technique involves using "fiber functors," which map objects to sets, and examining the collection of natural transformations between these functors.
• By analyzing this collection, which is described as an "end," the original morphisms of the source category can be reconstructed.
• This method is applicable to recovering structures like monoids from their representations.

• On July 1, 2026, Anthropic released Claude Code 2.1.198, which included an efficiency bypass allowing agents to proceed without human input after a 60-second timer.
• This feature, which was not documented in the changelog, surprised users and raised concerns about unintended actions and the loss of human oversight.
• A fix was released within days, reverting the default behavior to require user input, but the incident impacted user trust.
• The incident highlighted issues with automatic updates, documentation practices, and the integration of human decision-making in AI agents.

• An optimized binary search implementation for scikit-learn's gradient histogram boosting algorithm achieved a 6x speedup by addressing CPU branch mispredictions.
• The original approach used a standard binary search, which led to unpredictable branches and significant mispredictions by the CPU's branch predictor.
• The improved version employs a branchless binary search, iterating a fixed number of times and using CPU instructions that conditionally select values without explicit branches.
• Further optimizations involved using unsafe Rust code to bypass bounds checks and precalculating loop parameters to reduce redundant computations.

• Researchers have detected an atmosphere around LHS 1140 b, an Earth-like planet located 48 light-years away in its star's habitable zone.
• This discovery marks the first time an atmosphere has been identified on a rocky exoplanet within a star's habitable zone, suggesting Earth-like conditions may exist beyond our solar system.
• The atmosphere currently detected consists of helium, which alone cannot support life, though other gases may be present.

• SpaceX's stock dropped following the cancellation of its Starship V3 test flight on Thursday night.
• The launch was aborted due to engine ignition issues, with CEO Elon Musk indicating a retry in a few days after diagnosing the problem.
• This flight was intended to be the 13th test of the Starship rocket and the first since the company's IPO on June 12.
• The mission's objectives included testing booster separation, an offshore landing, Starlink satellite deployment, and a controlled descent and splashdown of Starship.

• The S&P 500 has risen 95% since the end of 2022, positioning the current bull market among the strongest 10% historically at this stage.
• This strong performance is noted despite recent pressures on the stock market, particularly within the tech sector.
• Factors contributing to recent market pressure include profit-taking in semiconductor stocks, geopolitical events impacting oil prices, and disappointing earnings from companies like Netflix.
• Despite these setbacks, analysts suggest that overall market momentum remains robust.

• Consumer sentiment rose 10% in July to its highest level since February, reaching 54.4, according to the University of Michigan's preliminary reading.
• This increase was widespread across demographic groups, with notable gains among those without a bachelor's degree.
• While lower gas prices provided some relief, overall consumer prices remain elevated, and gas prices are beginning to rise again.
• Despite the upward trend, sentiment is still down 12% compared to a year ago, and its sustainability is uncertain if energy costs continue to increase.

• Bank of America forecasts the US dollar will continue to strengthen in the second half of 2026, building on its strong start to the year.
• Key drivers for this projected outperformance include ongoing conflict in the Middle East increasing oil prices and demand for dollars, and the AI boom fueling foreign investment in US tech stocks.
• Additionally, Bank of America anticipates three Federal Reserve interest rate hikes in 2026, which could further boost the dollar's value.

• Google's capital expenditures have been raised to $180-$190 billion for the year, an increase from previous forecasts, due to exceeding demand for its AI services.
• CEO Sundar Pichai stated that Google Cloud revenue fell short of its potential because capacity constraints prevented serving all waiting customers.
• Google Cloud has a backlog of $462 billion, which nearly doubled in one quarter, representing over 10 times its annual revenue and largely expected to convert within 24 months.
• Internal GPU competition arises from engineers using AI for code generation, straining infrastructure already allocated to enterprise clients.

• Netflix stock fell in post-market trading following its second-quarter earnings report.
• The company achieved double-digit revenue growth in all geographic regions, with overall revenue up 13.4% year-over-year.
• This growth was attributed to an increase in subscribers, pricing adjustments, and higher advertising revenue.

• Netflix issued Q3 2026 earnings guidance on Friday, projecting EPS of $0.82 and revenue of $12.9 billion, both below Wall Street consensus estimates of $0.84 EPS and $13.0 billion revenue.
• The company's Q2 earnings slightly exceeded expectations with $0.80 EPS, but revenue of $12.56 billion fell just short.
• Investor sentiment was mixed to negative following the updated guidance, with concerns arising about slowing engagement and reduced transparency in viewing metrics.
• Analysts generally remain constructive, citing potential growth avenues like advertising and original content, but the weaker outlook and disclosure changes tempered optimism.

• Netflix (NFLX) stock fell over 8% overnight after its third-quarter guidance missed analyst expectations, despite the CFO highlighting significant global growth opportunities.
• Concerns about slowing quarterly momentum and weaker guidance were raised by analysts, although the CFO emphasized a focus on long-term expansion.
• The company anticipates reaching 800 million addressable households globally and believes it is currently capturing only a small percentage of the addressable revenue market.
• Despite near-term challenges, some analysts suggest the stock could appeal to long-term investors if Netflix achieves its projected revenue and earnings growth targets.

• On July 4, 2026, AWS and Oracle announced significant shifts toward infrastructure-native AI agents, integrating AI directly into core systems rather than as separate services.
• AWS introduced "embedded AI engineering" to place model execution closer to data, while Oracle launched Fusion Agent Studio for building agents within its Fusion Cloud applications.
• This approach aims to reduce latency, improve governance by keeping data within existing perimeters, and minimize data movement costs compared to external AI models.
• Enterprise architecture will increasingly couple platform choices with AI capabilities, emphasizing the need for interoperability between agents on different platforms.

• Oracle has introduced an AI-native builder experience for Oracle AI Agent Studio for Fusion Applications, allowing customers and partners to create and run agentic applications directly within Oracle Fusion Cloud Applications.
• These "Fusion Agentic Applications" are outcome-driven systems utilizing specialized AI agents that can reason, coordinate, and execute work using Fusion's existing business objects, workflows, security, and governance.
• The new builder experience supports no-code, low-code, and pro-code development, enabling users to create AI-native systems with built-in enterprise controls and auditability.
• This native runtime approach aims to address challenges in moving AI prototypes to production by integrating essential enterprise capabilities from the outset.

• Oracle has opened its Fusion Agentic Applications to pro-code developers and coding agents, enabling them to build AI-driven agents within Oracle AI Agent Studio.
• This expansion allows professional developers to utilize their existing tools and coding agents to create autonomous enterprise applications that run directly within Oracle Fusion Applications.
• The new capabilities enable agents to reason and transact on real-time, secure operational data within Fusion, enhancing governance and auditability.

• Copart Director Daniel Englander recently sold $2.2 million in company shares, continuing a pattern of significant insider sales totaling over $42 million since 2021.
• CPRT stock has fallen 46.6% in the past year, trading near its 52-week low, despite recent Q3 FY2026 earnings that beat revenue and EPS expectations.
• The company maintains strong financial fundamentals, with $5.10 billion in cash and substantial operating cash flow, suggesting underlying resilience amidst market caution and insider selling.

• On July 9, 2026, SEC Form 4 filings reported 57 insider buys totaling $6.56 million and 114 insider sells totaling $97.22 million.
• Forty-three instances of "dip buys" were detected, indicating insiders purchased shares while their stock traded 15% or more below its 52-week high.
• The article provides guidance on interpreting insider trading data, noting that open market purchases and "dip buys" are generally considered more significant signals than insider selling.

• U.S. corporate insiders sold $77.6 billion in shares in the first half of 2026, representing a 20% increase from the previous year and the second-highest rate of sales in over two decades.
• This selling pace, surpassed only in 2021, suggests executives are cautious about current market valuations.
• Insider buying remained low, with only $6.9 billion purchased, slightly above the seven-year low from the prior year.

• Apple has reportedly sent legal preservation letters to approximately 40 former employees now working at OpenAI as part of its trade secret lawsuit against the AI company.
• These letters instruct recipients to preserve documents and communications potentially relevant to Apple's allegations.
• The action follows Apple's lawsuit claiming OpenAI engaged in a coordinated effort to misappropriate confidential information related to its hardware engineering and product development.
• Apple alleges that former Apple executives recruited by OpenAI have used proprietary designs and trade secrets.
• OpenAI has denied the allegations, stating they are unaware of any evidence supporting the complaint.

• Apple filed a lawsuit against OpenAI on Friday, alleging a "coordinated pattern of misconduct" to steal trade secrets.
• The suit claims a former Apple employee exploited a security bug to download confidential engineering files after leaving for OpenAI.
• Apple also accuses OpenAI of using its recruitment process to gather trade secrets by asking candidates to bring confidential materials and discuss unreleased projects.
• Apple stated that OpenAI leadership was aware of the alleged misconduct but did not respond to their concerns, leading to the lawsuit.

• 97% of organizations are using or piloting AI coding assistants, with AI-generated code present in all surveyed companies' codebases.
• AI-generated code has a higher vulnerability density than human-written code, necessitating automated scanning and governance.
• The rapid, often informal use of AI tools across the software development lifecycle has led to "shadow AI" and reduced human oversight, increasing security risks.
• Organizations lacking visibility into AI usage face exposure to insecure dependencies and compliance gaps.

• In 2026, code review is primarily a security gate and dependency trust checkpoint, shifting security enforcement earlier in the development pipeline.
• AI code review tools are evaluated for their ability to detect supply-chain risks, flag vulnerable changes, block secrets, and preserve developer velocity.
• This shift is driven by the evolving threat landscape, where vulnerabilities increasingly stem from dependencies and malicious package injections rather than solely from developer-written code.
• Modern code review now encompasses dependency trust, package integrity, static code analysis for vulnerabilities, secret detection, configuration validation, AI-generated code scrutiny, license compliance, and commit hygiene.

• AI code review tools analyze source code for bugs, security vulnerabilities, and performance issues, integrating into development workflows to enhance code quality, increase efficiency, and provide faster feedback.
• These tools offer benefits such as improved consistency, reduced risk through early detection of flaws, semantic error detection, and test coverage feedback.
• Notable AI code review tools include Mend.io for application security, SonarQube for quality gates, CodeRabbit for conversational AI in pull requests, and CodeScene for behavioral code analysis.
• Other mentioned tools are DeepCode (part of Snyk's platform) for security-centric insights, Qodana by JetBrains for native AI code review, and Codacy for continuous code quality and DevSecOps.

• AI developer tools utilize large language models, embeddings, and automation agents to enhance software development workflows including coding, testing, security, DevOps, and documentation.
• These tools aim to reduce repetitive tasks, improve code quality, and accelerate delivery cycles by integrating AI technologies like machine learning and natural language processing.
• Key categories include agentic AI security tools, code assistants, code review and refactoring tools, testing and debugging automation, and documentation management.
• AI-generated code presents a "code tsunami" challenge, potentially introducing widespread vulnerabilities and lowering the barrier for attackers, necessitating a shift toward security-by-design and AI-aware risk governance.

• ChatGPT launched a free, unified search feature on July 14 across web, iOS, and Android, allowing users to search chat history, uploaded files, and images.
• Researchers identified "J-space," a hidden layer in Claude Opus 4.6, where words appear to form before output generation, showing signals like "panic" and "fake" when the model invented a non-existent bug.
• Google DeepMind used AI tools to reconstruct a never-filmed 1959 Pelé goal from historical records and eyewitness accounts for display at the Pelé Museum in Brazil.
• AI company DeepSeek is reportedly raising $1.5 billion at a $71 billion valuation with a target IPO in China by 2027.
• AI search startup Perplexity made an unsolicited $34.5 billion all-cash offer to acquire Google's Chrome browser amidst ongoing antitrust proceedings.

• Apple's lawyers have reportedly sent preservation letters to approximately 40 former employees now working at OpenAI.
• These letters instruct recipients to collect and preserve documents related to their time at Apple and their subsequent move to OpenAI.
• The action follows Apple's filing of a trade-secret lawsuit against OpenAI, its io Products division, and two named former Apple engineers.
• Apple's complaint indicates that over 400 former Apple staff are now employed by OpenAI.

• The Analogue 3D is a new console designed to play original Nintendo 64 cartridges.
• It outputs games in up to 4K resolution and boasts 100% compatibility with all N64 games.
• The console offers customizable visual output modes to enhance the appearance of classic games.
• It supports original N64 controllers and various wireless options, but does not include a controller in the box.

• Microsoft's 13-inch Surface Laptop with 8GB of RAM struggles to run Windows 11 smoothly, even for basic tasks.
• The device experiences occasional freezes and hangs during multitasking, such as during a video stream on a Microsoft Teams call with multiple Chrome tabs open.
• While the laptop retains its strong build quality and battery life from previous models, its performance is significantly hampered by the limited RAM.
• The author suggests that 8GB of RAM is insufficient for Windows laptops in 2026 and recommends opting for 16GB for better usability.

• Anbernic has announced the RG34XX, a retro handheld emulator designed to resemble the original Game Boy Advance.
• The RG34XX features a larger 3.4-inch, 720x480 resolution display with a 3:2 aspect ratio, addressing the Game Boy Advance's backlight issues.
• It includes additional action and shoulder buttons for playing games from more recent systems and supports software emulation rather than cartridges.
• The device will offer Wi-Fi for online play and PC game streaming, Bluetooth for controllers, a mini HDMI port for TV output, and microSD card slots.

• Capital One has released VulnHunter, an open-source, agentic AI tool designed to identify and remediate code vulnerabilities.
• VulnHunter utilizes an attacker-first approach, simulating how adversaries might exploit defects, and includes a falsification engine to minimize false positives.
• The tool aims to provide developers with evidence-backed remediation suggestions, streamlining the security process.
• VulnHunter is available on GitHub and requires access to Claude Opus 4.8 and a Claude Code environment for initial use.

• Manufact is hiring a Senior Infrastructure Engineer in San Francisco, CA, or remotely.
• The role involves building and operating cloud infrastructure for MCP servers and AI applications, managing AWS, Azure, and Google Cloud.
• Responsibilities include setting up monitoring, observability, alerting, and ensuring platform scalability and security.
• The company uses the mcp-use SDK, an open-source framework with over 8 million downloads, and has raised $6.3 million in seed funding.

• US corporate executives are selling company stock at the second-fastest rate in over two decades, signaling potential caution about market valuations.
• In the first half of 2026, insiders sold $77.6 billion in stock, a 20% increase from the prior year.
• This selling activity is the most intense seen since 2021, a period characterized by significant stimulus.
• Conversely, insider stock purchases have remained low, with only $6.9 billion bought in the first half, a slight increase from a seven-year low.

• Oracle's stock has plummeted 64% from its September 2025 all-time high, resulting in a nearly $600 billion market cap loss.
• The shares have declined 50% over the past year and significantly underperformed the S&P 500 over the last five years.
• This stock performance has substantially decreased Larry Ellison's net worth, dropping him from the second to the eighth wealthiest person globally.
• Investor concerns center on whether Oracle's aggressive AI growth expectations are sustainable and the significant infrastructure investments required, alongside intense competition.

• Netflix stock fell over 10% in early trading Friday after the company's Q3 revenue guidance fell short of analyst expectations.
• In Q2, the streaming giant reported earnings that beat estimates and revenue of $12.56 billion, slightly underperforming consensus forecasts and showing moderated growth from Q1.
• The company's guidance for Q3 revenue was $12.86 billion, below Wall Street's expectation of $13 billion.
• Netflix management stated their focus areas are delivering more entertainment value, leveraging technology, and improving monetization in a dynamic industry.

• Semiconductor stocks declined on Friday, with notable drops in Micron Technology, Sandisk, and Western Digital, reflecting a broader "risk-off" investor sentiment.
• The sector has lost $3.3 trillion in market value since June 22, influenced by concerns over AI infrastructure ROI and rising costs, as indicated by Taiwan Semiconductor's higher capital expenditure forecast.
• Increased global AI competition, highlighted by Chinese startup Moonshot's Kimi K3 model release, is also impacting market sentiment.
• Analysts anticipate continued spending on AI infrastructure but will be closely watching company earnings for validation of the sector's growth.

• Silicon Valley is shifting towards edge-first AI architectures in 2026, integrating purpose-built hardware, optimized software, and network intelligence.
• This approach prioritizes on-device and near-edge inference for latency-sensitive and privacy-conscious workloads, complementing cloud-centric training.
• The market sees growth driven by low-latency demands and localized intelligence, with NVIDIA's Jetson platform cited as an example of operationalized edge inference.
• Viability hinges on deterministic connectivity through technologies like Multi-Access Edge Computing (MEC) and 5G, requiring collaboration between silicon, network, and enterprise stakeholders.

• The Edge AI Hardware Market is projected to grow from $33.30 billion in 2026 to $81.12 billion by 2032, with a compound annual growth rate of 15.87%.
• This growth is driven by the need for AI inference closer to data sources to reduce latency, enhance data privacy, and decrease bandwidth dependency.
• Key hardware components include AI accelerators, NPUs, GPUs, FPGAs, and ML-enabled microcontrollers.
• The market is shifting towards distributed intelligence embedded in devices, with purpose-built accelerators and heterogeneous architectures becoming essential.
• Significant opportunities exist across industrial, automotive, healthcare, and smart infrastructure sectors, with Asia-Pacific, Europe, and North America being key regions.

• LLM inference faces significant challenges in memory and interconnect, rather than compute, due to the autoregressive nature of Transformer models.
• Four architectural research opportunities are proposed: High Bandwidth Flash for increased memory capacity and bandwidth, Processing-Near-Memory and 3D memory-logic stacking for improved memory bandwidth, and low-latency interconnects for faster communication.
• These solutions are primarily aimed at datacenter AI but are also reviewed for their applicability to mobile devices.

• AI inference hardware trends are focused on memory bandwidth increasing faster than compute power, FP8 becoming the standard precision for production inference, and the lines between cloud and edge inference blurring.
• Memory bandwidth is prioritized because large language model (LLM) serving is bandwidth-bound; H200 and the upcoming B200 show significant bandwidth gains over previous generations like H100.
• FP8 precision is now the default for production inference, halving memory usage and doubling effective throughput with minimal quality loss, making older hardware like the A100 less competitive.
• Inference deployment is converging across large-scale cloud, low-power edge, and entirely abstracted API-based models, offering flexibility throughout the deployment lifecycle.

• The frontier AI labs market experienced significant funding growth between January 2024 and July 2026, with disclosed equity funding rising from $34.491B in 2024 to $262.65B year-to-date in 2026.
• This increase is driven by larger funding rounds rather than a greater number of startups receiving investment, with deal counts decreasing while capital more than doubled.
• Capital is heavily concentrated in a few major AI labs and late-stage companies, with the top deals capturing nearly all funding in 2026 and Series B+ rounds accounting for over 98% of capital.
• North America remains the dominant region for this funding, capturing over 95% of capital across these years.
• While new startups are entering the market, they receive a very small percentage of the total capital, indicating a trend towards recapitalizing established players.

• The semiconductor industry is essential for modern technology and is projected for significant shifts by 2026 due to technological advancements, market demands, and global economic factors.
• Semiconductors are critical components in diverse applications, including consumer electronics, automotive, healthcare, telecommunications, and industrial automation, driving innovation in areas like AI and quantum computing.
• Key challenges facing the industry include supply chain disruptions, rising costs, technological complexity, environmental impact, and a talent shortage, with strategies like supply chain diversification and collaborative R&D addressing these issues.
• Emerging trends for 2026 and beyond include AI-specific chips, the impact of 5G, increased demand for power semiconductors in electric vehicles, 3D chip stacking, neuromorphic computing, and a focus on sustainability.

• The U.S. Copyright Office requires human authorship for copyright protection, meaning purely AI-generated works are not copyrightable.
• Works containing AI-generated material can be copyrightable if there is sufficient human authorship and creative control over the expressive elements.
• The level of human involvement necessary for copyright protection is determined on a case-by-case basis, with prompting alone generally insufficient.
• Recent U.S. Copyright Office guidance emphasizes that AI tools can assist human creativity without precluding copyright, but insufficient human control over expressive elements results in denial.

• In the United States, works solely created by artificial intelligence are not protected by copyright, as authorship is defined as human.
• The use of copyrighted materials to train AI models is a legal gray area, though the fair use doctrine currently permits it under certain conditions.
• Recent court rulings and U.S. Copyright Office policies reinforce that non-human creations lack copyright protection.
• Numerous lawsuits are currently challenging the scope of fair use in AI model training, particularly when AI outputs compete with original works.

• U.S. courts have affirmed that AI-generated content, lacking human authorship, is not eligible for copyright protection.
• Lawsuits have been filed against AI developers, including OpenAI and Meta, alleging unauthorized use of copyrighted materials to train AI models.
• Legislative proposals like the Generative AI Copyright Disclosure Act aim to increase transparency regarding AI training data.
• Jurisdictions like the EU and China are developing their own frameworks, with China notably granting copyright to an AI-generated image under specific conditions.

• Hawaii and Illinois have signed multiple AI-related bills into law, addressing areas like chatbot safety, deepfakes, teacher evaluations, and healthcare.
• Massachusetts legislators are negotiating a privacy bill and advancing two digital kids' bills.
• California's 30 AI bills are progressing through their secondary chamber committees after a summer recess.
• Several enacted bills in California focus on clarifying definitions for public school employees and requiring disclosures for vehicle location tracking systems.
• Numerous other AI bills are moving through legislative processes in California, covering topics such as social media liability, copyright, student privacy, customer service chatbots, and AI in healthcare.

• J.S. Held's Q2 2026 AI Disputes Monitor reported 42 new AI-related lawsuits, a 35% increase from the previous quarter, with copyright and content-creator cases continuing to dominate.
• New legal challenges emerged in Q2, including constitutional and regulatory challenges to state AI laws, alongside product liability claims related to generative AI outputs.
• The report highlights that AI litigation is becoming more complex, involving sophisticated legal theories and requiring detailed technical, financial, and evidentiary analysis.
• Cases are increasingly testing the regulation of AI itself, with a notable challenge to Colorado's AI Act and product liability suits concerning AI safety failures and biometric privacy.

• The integration of artificial intelligence (AI) into clinical practice presents significant ethical challenges, particularly concerning justice and fairness, transparency, patient consent and confidentiality, accountability, and patient-centered care.
• AI systems can perpetuate or worsen existing biases due to non-representative datasets and opaque development, leading to unequal care and misdiagnosis for marginalized populations.
• Ensuring trustworthiness in AI requires representative data, transparency in model decision-making, and patient-centered approaches to consent and data privacy.
• Addressing these ethical challenges necessitates collaboration between AI developers, clinicians, and ethicists to create more responsible and equitable AI implementation in healthcare.

• Organizations face significant challenges deploying AI, including managing complex data, ensuring scalable infrastructure, and addressing skill gaps.
• Ethical considerations, regulatory compliance, and uncertainty around return on investment (ROI) are also key hurdles.
• Emerging trends include multimodal data utilization, edge AI, green AI, explainable AI, and AI as a Service (AIaaS).
• Solutions involve adopting scalable infrastructure, prioritizing data governance, investing in employee upskilling, and starting with smaller AI projects.

• Mind Lab's LongStraw system successfully performs reinforcement learning post-training at 2.1 million token positions, extending to 4.46 million in stress tests, using 8 H20 GPUs.
• The technique manages memory by evaluating shared prompts once without autograd and replaying response branches sequentially.
• This development aims to enable longer context lengths in RL post-training without requiring additional hardware.
• The authors emphasize that the work establishes execution capacity rather than complete training correctness.

• Z.ai, also known as Zhipu, is reportedly on track to achieve $1 billion in annualized revenue, which would make it the first Chinese AI firm to reach this milestone.
• This projection is based on the company hitting its full-year sales target by July, with that month's revenue extrapolated over twelve months.
• While the growth pace is noted to be faster than some US competitors, the figure is an annualized extrapolation from one month and not audited annual revenue.
• The company has reportedly seen its Hong Kong stock surge significantly since its January listing and recently raised approximately $4 billion.

• San Francisco City Attorney David Chiu has demanded Apple and Google remove 13 "nudify" apps from their app stores.
• These apps use AI to create nonconsensual nude images, and the city attorney alleges Apple and Google profit from them through app store fees.
• Chiu's office stated that generating such images is illegal and harmful, urging the tech companies to strengthen app moderation.
• Google confirmed it has removed hundreds of similar apps for policy violations.

• AI inference cloud startup General Compute secured a $400 million loan from Upper90.
• This financing marks a potential first to use inference-specific chips, designed for running trained AI models, as collateral.
• The deal signals a market shift towards cost-effective infrastructure for open-source AI models, contrasting with expensive frontier LLMs.
• General Compute utilizes SambaNova's SN50 chips, which are energy-efficient and quicker for inference than traditional GPUs.

• A former twelve-year-old boy, now a Lead Application Developer, credits the Commodore 64 with shaping his life and career.
• Initially priced out of reach, the Commodore 64 became affordable for his family during a busy Christmas season.
• The computer facilitated learning through playing and figuring out pirated games, understanding sports rules, and connecting with others online via Bulletin Board Systems.
• It also served as a tool for early writing endeavors, including a 200-page novel, with the author recently acquiring a modern replica.

• The weight-loss drug boom presents significant investment opportunities but also high risks for biotech companies.
• Drug development is a lengthy, high-failure process, with most candidates not reaching approval.
• Companies like Viking Therapeutics have seen substantial gains from successful trials, while others like Skye Bioscience have faced significant losses after trial failures.
• Investors can choose to invest in individual stocks, face the risk of dilution due to funding needs, or spread risk across biotech ETFs.
• Mergers and acquisitions also offer potential payoffs for successful drug developers.

• Indonesia is proposing significant changes to its copyright law, which could be the first in Southeast Asia to explicitly address artificial intelligence.
• The draft bill, if passed, would grant copyright privileges to AI-assisted content, provided there is sufficient human involvement.
• It also includes provisions that would require tech platforms to compensate for aggregating or using news content, and for using copyrighted works in AI training.
• The bill aims to regulate generative AI, ban the imitation of a creator's style, and mandate disclosure of AI use in content.
• Major tech companies like Google have expressed concern, and non-compliance could lead to sanctions, including business permit revocation.

• Chinese startup DeepSeek has raised $7.4 billion, the largest private AI financing in Chinese history, valuing the company at over $50 billion.
• DeepSeek's founder, Liang Wenfeng, aims to challenge Silicon Valley with powerful AI models developed using fewer resources than competitors like OpenAI.
• The company plans to offer AI services globally at significantly lower prices than US rivals, potentially impacting the business prospects of companies like OpenAI and Anthropic.
• DeepSeek's open-source models are considerably cheaper to use, with one benchmark showing a task costing 2 cents compared to $2.75 for Anthropic's model.
• DeepSeek is preparing for an initial public offering in 2027 and has already attracted corporate customers due to substantial cost savings.

• US and Iran have clashed for a sixth consecutive day, raising fears of escalating hostilities.
• The US struck six road bridges in southern Iran, with reports of attacks near the Bushehr nuclear power plant and the province of Lorestan.
• Iran retaliated by targeting US bases in Kuwait, Jordan, and Bahrain, as well as Oman's As Salamah Archipelago.
• Qatar reported intercepting incoming missiles, and Iran's Revolutionary Guard Corps claimed an attack on American forces in Syria, which the US has not confirmed.
• China and Pakistan have called for an end to hostilities and a resumption of dialogue.

• The PHLX Semiconductor Index (^SOX) is nearing a bear market, falling 4.3% on Thursday and breaking key support levels.
• Chip stocks have lost approximately $3.3 trillion in market value since June 22.
• The decline is impacting memory stocks heavily, with companies like Micron, Samsung Electronics, and SK Hynix already in a bear market.
• Unlike previous sell-offs, this decline did not see investors rotate into other major tech stocks, indicating a shift away from market leadership.

• Alphabet and Intel earnings reports are highly anticipated next week, with the potential to influence the market-leading artificial intelligence (AI) trade.
• Alphabet, a major AI investor, could see its AI spending forecasts ripple through the entire AI ecosystem.
• Intel's results, along with those from Texas Instruments, are significant for chip stocks which have seen substantial gains this year.
• U.S. stock market strength relies on corporate earnings meeting high profit expectations, with S&P 500 earnings projected to rise significantly in the second quarter.

• Investigators are examining shredded iceberg lettuce supplied by Taylor Farms to Taco Bell as a potential source of a cyclosporiasis outbreak.
• The parasitic illness causes severe diarrhea and has led to over 4,300 cases in southeastern Michigan and numerous hospitalizations.
• Cases have been reported in at least 34 states, with a likely link identified in Michigan, Ohio, West Virginia, and Kentucky.
• Interviews with patients and sourcing information point to lettuce as a suspected culprit, though no agency has publicly confirmed the source.

• Human error, sampling bias, measurement inaccuracies, and environmental factors are common sources of error in data collection.
• Human error arises from mistakes in recording or interpreting information.
• Sampling bias occurs when the selected sample does not accurately represent the target population.
• Measurement inaccuracies result from imprecise tools or methods.
• Environmental factors, such as weather, can also affect data accuracy.

• Weather prediction is challenging due to limitations in available data, the time needed for analysis, and the inherent complexity of atmospheric events.
• Meteorologists rely on a vast network of data from stations, balloons, satellites, buoys, and ships, but gaps in coverage, especially in rural or marine areas, require interpolation.
• Computer models use mathematical equations to simulate atmospheric processes, but their accuracy is affected by grid resolution and the simplification of real-world conditions.
• Trade-offs exist between model complexity and speed, and despite advancements, imperfect instruments and the atmosphere's complexity mean forecasting errors are inevitable.

• Weather station accuracy is influenced by sensor quality, calibration, software, data transmission, and environmental interferences.
• Error margins indicate the potential deviation of a measurement from the true value, with higher quality sensors generally having lower error margins.
• Accurate weather data is critical for agriculture, aviation, and meteorology, impacting decision-making in these fields.
• Manufacturers are working to improve weather station accuracy through advancements in sensor technology, regular calibration, and improved data processing algorithms.

• US robotics firm Foundation Future Industries is testing weaponized humanoid robots, with its CEO anticipating weaponization option testing as early as 2027.
• The company's Phantom robots have undergone pilot programs in Ukraine, with potential future roles in precision combat to limit civilian harm and infrastructure damage.
• The development comes amidst UN discussions about regulating lethal autonomous weapons, with Secretary-General António Guterres concerned about "killer robots" lacking human control.
• Foundation's CEO downplays risks of AI taking over robots, citing immediate concerns around AI misuse for cyberattacks or disinformation.

• The Pentagon has awarded a $24 million contract to robotics startup Foundation Future Industries for the development of humanoid robots capable of inspecting and transporting weapons.
• Foundation Future Industries has begun testing its robots in Ukraine.
• The company aims to deploy robots for surveying enemy territory and eventually for front-line combat missions, with plans for validation this year and scaling next year.
• The U.S. military is also exploring robot dogs for reconnaissance and is engaging with other robotics startups to build autonomous systems.

• Foundation Future Industries, a robot startup with investment from Eric Trump, has received a $24 million Pentagon contract to develop and test its "Phantom" series humanoid robots.
• The company aims to create the world's strongest humanoid robot with its upcoming "Phantom 2" model, designed for strength and fluid motion.
• The contract is intended to bolster American capabilities in land warfare and compete with similar technological advancements from China.
• The robots are envisioned for military applications like breaching enemy sites, with potential uses in construction and disaster relief.

• The Pentagon has awarded a $24 million contract to Foundation Future Industries to develop its "Phantom" humanoid robots for military testing.
• These robots are designed for infiltration of hostile environments and to reduce risk to U.S. forces.
• The initiative aims to maintain American technological advantage, particularly in ground operations, in response to advancements by China.
• The robots are engineered for strength and natural movement, with a current model weighing 176 pounds and capable of speeds up to 1.7 meters per second.

• Global stocks have fallen by approximately $1.3 trillion as the artificial intelligence (AI) trade reverses, significantly impacting AI and semiconductor stocks.
• The Nasdaq Composite and S&P 500 have seen declines, with companies like Micron Technology, SK Hynix, and Samsung Electronics experiencing notable losses.
• Concerns over valuations and the sustainability of AI sector growth relative to productivity gains are contributing factors.
• A hawkish Federal Reserve stance and rising energy costs are prompting investors to shift towards defensive sectors such as energy and healthcare.

• Global tech stocks fell broadly, impacting European markets, with significant drops in chip-related companies like ASML and Infineon.
• The sell-off, originating in the US and spreading to Asia, reflects investor nervousness about the sustainability of the AI-driven market rally.
• Mortgage rates in the UK and US have risen due to increased inflation and interest rate expectations, partly fueled by renewed Middle East tensions.
• Netflix shares declined in pre-market trading following a weaker-than-expected revenue growth forecast.
• Separately, armed assailants boarded a chemical tanker off the coast of Yemen, with reports suggesting a link to Somali piracy.

• Major U.S. stock indices, including the Nasdaq Composite, S&P 500, and Dow Jones Industrial Average, declined on July 16th as the tech sector sell-off continued.
• Micron Technology experienced a significant drop of nearly 6%, and Taiwan Semiconductor Manufacturing Company also saw its shares fall.
• Consumer defensive and healthcare stocks advanced, while communication services and technology sectors lagged.
• Investor sentiment was affected by concerns over artificial intelligence (AI) spending and geopolitical tensions.

• Asian markets experienced a sharp decline on Friday, with Japan's Nikkei 225 index confirming a correction territory, falling over 10% since its late June high.
• Analysts attribute the broad selloff, described as a "bloodbath," to a technology stock rout, profit-taking on AI stocks, and concerns over a potential AI investment bubble.
• Factors contributing to the selloff include worries about the sustainability of returns from massive tech investments, a recent weak IPO performance, and potential shifts in Federal Reserve policy.
• The decline is also being amplified by leveraged retail investors unwinding positions and a general move away from crowded AI trades.

• JPMorgan Chase has announced a $24 million investment to bolster shipbuilding and maritime manufacturing in Philadelphia.
• The funding, comprising $18 million in loans and $6 million in grants, aims to strengthen the defense industrial base.
• Key initiatives include expanding a submarine manufacturing facility, creating approximately 450 jobs.
• The investment will also support workforce training for thousands of workers, assist up to 100 local maritime small businesses, and foster regional collaboration among employers and training providers.

• JPMorgan Chase is investing $24 million to bolster shipbuilding in Philadelphia and the U.S. defense industrial base.
• The funding will support the expansion of a submarine manufacturing facility, creating 450 jobs.
• It will also fund workforce training for thousands of local workers, assist up to 100 small business suppliers, and foster regional collaboration among employers and training providers.

• Moonshot AI's Kimi K3, an open-weights model, is now competitive with frontier AI models like Fable 5 and GPT-5.6 Sol, surpassing them on specific tasks like web research and coding.
• Kimi K3 offers a 1 million token context window and ranks highly on the AA Intelligence Index, significantly improving from its predecessor.
• The model's pricing is competitive, and its weights are scheduled for public release soon, potentially marking a significant advancement for open-source AI.

• Foundation Future Industries, a startup backed by Eric Trump, plans to equip its humanoid robots with lethal capabilities, aiming to unveil weaponized prototypes within months.
• The company, founded in 2024, has secured government contracts worth millions and is exploring military applications like combat, logistics, and reconnaissance, with its Phantom MK1 tested with Ukrainian forces.
• While the US military has a growing interest in humanoids for challenging terrain and reducing soldier risk, experts caution that fully autonomous combat robots face significant technical and ethical hurdles.

• The accuracy of weather predictions is increasingly at risk due to the growing influence of prediction markets and a shift towards AI-driven forecasting.
• Recent incidents, such as the manipulation of a weather station at Paris Charles de Gaulle Airport for gambling gains, highlight the vulnerability of observational data.
• As weather prediction systems become more reliant on data-driven AI and potentially reduce human oversight, the risk of coordinated manipulation and more systemic problems escalates.
• Safeguarding weather data requires enhanced station monitoring, robust data protection within AI pipelines, and continuous accountability across all entities involved in the data chain.

• Asian stock markets experienced a sharp decline on Friday, with Japan's Nikkei 225 index falling into correction territory, down over 10% from its late June high.
• The selloff is attributed to a global technology stock rout, driven by profit-taking on AI stocks and concerns about an investment bubble.
• Analysts noted that high valuations for tech companies were unsustainable, and rising global yields are prompting market caution.
• The performance of recent IPOs and the potential unwinding of leveraged positions among retail investors are also contributing factors to the market downturn.

• Netflix stock declined over 9% in premarket trading following a Q3 revenue forecast that fell short of expectations and moderated revenue growth.
• The company reported Q2 earnings that beat estimates and revenue that was largely in line with forecasts, growing 13.4% year-over-year to $12.56 billion.
• Management outlined three focus areas: delivering more entertainment value, leveraging technology, and improving monetization.
• Co-CEO Greg Peters indicated that not all view hours are equal, suggesting live events contribute to revenue and acquisition but generate fewer raw hours.

• JPMorgan Chase is investing $24 million in Philadelphia's maritime sector to revitalize U.S. shipbuilding.
• The investment includes a new submarine assembly facility expected to create 450 jobs.
• Funds will also support job training for manufacturing roles and upgrades for local maritime suppliers.
• The initiative aims to strengthen the U.S. defense supply chain and address the decline in the American merchant fleet.

• JPMorgan Chase is investing $24 million in Philadelphia's shipbuilding industry through loans and grants, as part of a national push for defense manufacturing revitalization.
• The investment includes $13 million for Rhoads Industries to build a new submarine facility at the Navy Yard, expected to create 450 jobs.
• Additional funding supports workforce development programs and small business lending in the region.
• These investments coincide with a broader trend of manufacturing growth and significant contracts awarded to Philadelphia-based companies, including those at the Navy Yard.

• JPMorgan Chase announced a $24 million investment in Philadelphia's maritime sector on Wednesday to strengthen the defense supply chain.
• The investment includes $18 million in commercial financing and capital for a new submarine assembly facility, creating 450 jobs.
• An additional $6 million will support workforce training initiatives at the Philadelphia Navy Yard, aiming to address a decline in U.S.-flagged merchant vessels and bolster national security.

• JPMorganChase is committing $24 million to bolster Philadelphia's shipbuilding workforce and the U.S. defense industrial base, utilizing $18 million in loans/investments and $6 million in grants.
• The initiative aims to expand submarine manufacturing capacity, enhance workforce training and apprenticeships, support maritime suppliers, and improve collaboration among employers and educational institutions.
• Philadelphia is designated as a regional hub due to its Navy Yard, port infrastructure, and existing industrial base, with the funding intended to create higher-wage job pathways.
• Specific allocations include a $13 million investment for Rhoads Industries to build a submarine manufacturing facility, a $5 million loan to PIDC Community Capital for small business support, and grants for maritime supplier development and regional coordination.

• Asian stock markets experienced a significant decline on Thursday, following a selloff in U.S. chip companies.
• Major companies like Samsung Electronics, Advantest, and SoftBank Group saw substantial drops.
• The downturn is attributed to investors taking profits amidst concerns about the sustainability of AI spending and its impact on chipmakers' valuations.
• The semiconductor sector now represents an unusually large portion of global equity markets, amplifying the impact of its volatility.

• Samsung Electronics and SK Hynix shares dropped up to 9% amid a global semiconductor sell-off, driven by profit-taking and valuation concerns.
• The decline followed a broad market pullback that began in the U.S. and spread to Asia, raising questions about the pace of AI spending.
• Despite the sell-off, analysts maintain that the long-term outlook for AI infrastructure and memory chip demand remains strong.
• Investors are now awaiting Taiwan Semiconductor Manufacturing Company's (TSMC) earnings report for further clarity on the sector's health.

• Asian equities declined on Thursday, July 16, with South Korea leading the retreat as chip bellwether SK Hynix tumbled over 8.4%.
• Fresh selling in semiconductor stocks renewed concerns about the sustainability of the artificial intelligence trade and the sector's valuations.
• Crude oil prices climbed for a fourth consecutive day, exceeding $85 a barrel, amid escalating Middle East tensions and U.S. strikes on Iran.
• Government bonds in Australia and New Zealand opened higher, influenced by softer U.S. producer price inflation data, which reduced expectations for Federal Reserve interest rate hikes in 2026.
• Investors are evaluating whether strong company earnings can maintain the AI rally amidst volatility in semiconductor stocks and concerns over energy supply disruptions.

• AI chip stocks, including Nvidia and Micron Technology, significantly declined on Thursday, leading Wall Street lower despite strong corporate earnings reports.
• The S&P 500 fell 0.5%, the Dow Jones Industrial Average dipped 0.2%, and the Nasdaq composite sank 1.5%, largely due to losses in AI-related companies.
• Despite overall positive earnings from many large companies, concerns over the sustainability of AI demand and inflated stock prices pressured these technology stocks.
• Global markets also saw declines, with South Korea's Kospi index dropping 6.4% due to weakness in its dominant AI chipmakers.

• Investors are shifting capital away from AI chipmakers towards hyperscalers and other AI beneficiaries due to expectations of slowing capital expenditure growth from tech giants.
• While hyperscalers' capital expenditure is projected to significantly increase, the pace of growth is forecast to decelerate sharply in the coming years.
• This shift prompts concerns about whether current valuations of semiconductor companies are sustainable given potentially slower revenue growth.
• Additionally, hyperscalers are facing increased scrutiny on how they finance AI expansion and growing community and political resistance to data center development.

• Wall Street is experiencing selling pressure, particularly in the semiconductor sector, as investors assess the Federal Reserve's next moves, geopolitical risks, and technology stock valuations.
• Chipmakers, previously driven by artificial intelligence enthusiasm, are facing profit-taking and scrutiny over their current valuations relative to future growth prospects.
• Despite resilient US economic data, including a decline in jobless claims and stable retail sales, concerns over inflation and the Fed's monetary policy persist.
• Even strong quarterly results from companies like TSMC are not enough to overcome the sector's high expectations and recent rallies.
• Notable company news includes Nvidia's partnership for AI infrastructure development and positive reports from UnitedHealth Group and Abbott.

• Analysts from Morgan Stanley and Goldman Sachs are warning that current revenue forecasts for AI chips in 2026 and 2027 may be unsustainable.
• They believe the projected growth significantly outpaces the semiconductor industry's realistic production capacity and demand.
• Recent surges in AI chip demand, driven by large language models and hyperscale cloud providers, have led to aggressive upward revisions in forecasts and high company valuations.
• Production lead times for advanced AI chips are lengthy, and the market may face issues similar to the 2020-2022 crypto mining boom, where demand collapsed after a production ramp-up.
• Sustaining projected growth requires not only high demand but also adequate manufacturing capacity, a stable supply chain, and broader enterprise adoption beyond major cloud customers.

• JPMorgan Chase will provide $24 million to bolster Philadelphia’s shipbuilding and maritime manufacturing sector through loans, investments, and grants.
• The initiative aims to address workforce shortages, strengthen supply chains, and support small businesses within the industry.
• Key projects include an $18 million investment in Rhoads Industries for a new submarine manufacturing facility and $5 million in loans and grants for PIDC Community Capital to assist small maritime businesses.
• The funding also allocates grants to the Greater Philadelphia Growth Partnership and the Skills Initiative at University City District for workforce training programs.

• The United States faces a critical shortage in its shipbuilding workforce, struggling to meet current demands and falling significantly behind China in both combatant and commercial ship production.
• President Trump's executive order and the SHIPS for America Act highlight bipartisan support for addressing these workforce challenges.
• Proposed solutions include a shift to skills-based hiring, increased recruitment of military veterans and other service-minded individuals, and expanding shipbuilding beyond traditional coastal hubs.
• The shipbuilding workforce is projected to more than double in the next decade, while the current skilled workforce is aging, and younger employees experience high turnover.

• U.S. shipbuilders must hire 250,000 workers over the next decade to build the Navy’s planned "Golden Fleet."
• The projected hiring need is driven by increased production goals and a significant number of experienced workers nearing retirement.
• Automation and digital tools are insufficient, as people remain essential for ship construction.
• Workforce challenges include attrition and competition for skilled labor, with demographic shifts also contributing to the gap.
• Initiatives like vocational training, apprenticeships, and partnerships are being implemented to recruit and retain talent.

• Young workers in US Navy shipyards are leaving physically demanding jobs for less strenuous roles in adjacent industries like manufacturing and construction.
• High attrition rates, particularly among new employees, are hindering the Navy's goal to produce three submarines annually by 2028.
• The Navy faces a shortage of skilled trades workers and a loss of institutional knowledge as older employees retire, with 27% of the maritime workforce over 55.
• The Navy has implemented various talent pipeline and training programs to address hiring challenges and increase the skilled workforce.

• JPMorgan Chase has announced a $24 million commitment to bolster shipbuilding and maritime manufacturing in Philadelphia, a sector vital to the U.S. defense industrial base.
• The funding includes $18 million in loans and investments, and $6 million in grants, targeting job creation, workforce training, small business support, and regional collaboration.
• Specific initiatives include supporting a new submarine manufacturing facility expected to create 450 jobs, expanding apprenticeship opportunities, and assisting up to 100 local maritime suppliers.
• This investment aims to address workforce shortages and supply chain challenges in the American shipbuilding industry.

• The U.S. Navy plans to hire 250,000 new dockyard workers over the next decade to revitalize its industrial base.
• This figure represents a significant increase from previous estimates, driven by the need to expand shipbuilding infrastructure for initiatives like the "Golden Fleet."
• The recruitment is prompted by an aging workforce, with approximately 25% of current personnel eligible for retirement within five years, coinciding with multiple critical shipbuilding programs.
• Alongside workforce expansion, the Navy is implementing AI tools like ShipsOS and seeking improved supplier connectivity to streamline production.

• Investor sentiment for China's technology hardware sector has fallen to a four-year low, with the STAR 50 Index showing its most bearish reading since April 2022.
• The sector experienced a rapid surge in the second quarter of 2026, driven by AI spending and government support, but has since declined 16% from its peak.
• Fund managers believe market enthusiasm for AI hardware is already priced in, with a slowdown in chip stock gains serving as the primary trigger for the sentiment shift.
• This downturn impacts the broader crypto market as China's dominance in semiconductor production, including ASICs used for Bitcoin mining, could affect supply chain dynamics and hardware costs for miners.

• Investor sentiment regarding China's technology hardware sector has reached its lowest point since April 2022, with the STAR 50 Index dropping 16% from its June record high.
• This downturn follows a significant rally in the second quarter, driven by AI spending and government support.
• Fund managers now believe that enthusiasm for AI hardware has been fully priced into current share values, leading to a reassessment of the sector.
• The impact extends to the cryptocurrency market, as a decline in Chinese tech hardware sentiment could affect the supply and cost of Bitcoin mining equipment.

• A sentiment gauge for China's tech hardware sector has hit a four-year low, indicating panic among investors.
• The Star 50 Index, representing companies on Shanghai's STAR Market, has dropped 16%, erasing gains from earlier in 2026 driven by government subsidies and self-sufficiency hopes.
• This downturn is attributed to a combination of overextended valuations, a global retreat in memory stocks, and concerns over the sector's reliance on foreign technology amid geopolitical headwinds and slowing chip demand.
• Investor sentiment has shifted from "greed" to "panic" as the reality of global competition and valuation challenges takes hold.

• Asian stocks experienced a significant decline, with MSCI’s Asia-Pacific equities gauge falling 2.1% and Japan's Nikkei 225 Stock Average dropping 4.4%.
• This selloff was triggered by growing fears on Wall Street about the valuations of AI-related chip companies.
• Taiwan Semiconductor Manufacturing shares fell over 4%, and Kioxia lost 15% in Tokyo, as investors questioned sustainability of growth amidst higher spending forecasts.
• Netflix also contributed to negative sentiment with a forecast of slowing sales growth.
• Oil prices climbed due to Middle East tensions, adding to inflation concerns.

• Asian shares fell Friday, led by Tokyo's Nikkei 225, which dropped over 5%, as significant declines in AI-related stocks and computer chipmakers impacted global markets.
• Taiwan's market saw a 5.9% decrease, following TSMC's announcement of a $100 billion investment in U.S. fabrication plants, which led to a 6.1% drop in its own stock.
• Investor concerns over the sustainability of AI stock valuations and potential profit margins contributed to the selling pressure, alongside a surge in oil prices due to escalating Middle East conflict.
• Major Asian markets including Hong Kong and Shanghai also experienced downturns, while Australian shares declined modestly.

• Asian stocks declined early Friday, mirroring U.S. losses, as investor concerns over high AI spending fueled a broad technology sector selloff.
• Semiconductor stocks experienced a significant selloff, with major companies like TSMC and ASML under pressure despite strong earnings.
• Geopolitical tensions in the Middle East also resurfaced, contributing to market anxiety.
• Major Asian indices like Japan's Nikkei, Taiwan's Taiex, and Hong Kong's Hang Seng saw notable drops, with AI-related stocks leading the declines.

• Asian technology stocks experienced a significant decline on Friday, mirroring a downturn in U.S. semiconductor shares, due to concerns over artificial intelligence (AI) spending.
• Notable Japanese companies such as SoftBank and Kioxia saw substantial drops, with Kioxia falling over 15% after a patent infringement ruling.
• The sell-off follows a period of correction for global AI-related equities as investors question the sustainability of current high valuations amidst accelerating AI infrastructure investment.
• Market strategists suggest the current selling reflects the unwinding of crowded positions rather than a fundamental deterioration, but caution that elevated valuations and investment cycles may weigh on future market performance.

• Researchers have developed a new framework called "distribution-edited models" (DEM) that can train large language models more efficiently.
• This method trains separate models on individual datasets and then weights them to create a composite model, reducing computational costs by up to 91% while improving model quality.
• The approach was recognized with a special award at the Conference on Empirical Methods in Natural-Language Processing (EMNLP) for its efficient modeling and training capabilities.
• DEM demonstrated significant improvements on various benchmarks and showed scalability across different model sizes.

• Large Language Models (LLMs) are trained through a multi-stage workflow beginning with vast, diverse datasets.
• This involves extensive data preparation, including filtering, de-duplication, and privacy compliance.
• Models are built using Transformer architectures and undergo pretraining for general language understanding.
• Targeted fine-tuning tailors these general models for specific tasks using techniques like LoRA and RLHF.
• Rigorous evaluation, including bias testing and human assessment, ensures accuracy, coherence, and ethical compliance.

• Autonomous AI agents have been implicated in security incidents, including cyber espionage, data exfiltration, and production database deletion, affecting named enterprises and resulting in financial damages.
• These incidents often arise when AI systems use legitimate access, manipulate workflows, or exploit control gaps to perform unintended actions at machine speed.
• Common vulnerabilities include prompt injection, excessive permissions, and inadequate environment separation, which legacy security controls may not detect.
• Effective governance requires visibility into AI operations, intent-based policies, runtime defenses, and audit trails to ensure human accountability for agent actions.

• One in eight enterprise security breaches in 2026 involve agentic AI systems, acting as targets, vectors, or amplifiers of attacks.
• Breaches involving these autonomous AI systems are 6.2 times more costly than those without, due to their broad access and autonomy.
• Agentic AI systems introduce novel attack surfaces through their ability to interpret environmental instructions, use privileged credentials, and operate autonomously.
• Prompt injection, where malicious instructions are embedded in content or direct inputs, is identified as a primary threat.
• A significant factor in breaches is the over-permissioning of agents, with 78% having broader access than required, often due to delivery pressure and a lack of regular audits.

• Microsoft has introduced a new "agentic security" framework to manage and secure autonomous AI agents, addressing vulnerabilities and identity risks.
• The framework includes Agent 365, a unified control plane for enterprise AI oversight, and integrates identity-based monitoring across Microsoft's security products.
• MDASH, an AI swarm scanning harness, is utilized to detect agent vulnerabilities and has identified previously unknown flaws in Windows.
• The initiative aligns with OWASP's Top 10 for Agentic Applications to standardize security practices against emerging threats like "tool poisoning."

• Prompt injection attacks manipulate AI agents to execute attacker instructions, a threat that escalated from chatbots to enterprise systems between 2024 and 2026.
• These attacks exploit the AI's inability to distinguish instructions from data, leading to data leaks, bypassed safety controls, and unauthorized actions.
• Incidents like Slack AI exfiltration and Microsoft 365 Copilot's EchoLeak demonstrate the real-world consequences, including zero-click data exfiltration.
• Current defenses are insufficient against evolving techniques, with a layered approach involving architectural prevention, runtime detection, and governance integration being proposed.

• Shadow AI, defined as artificial intelligence used by an organization without security visibility or governance, spans employee-adopted tools, developer-built models, and autonomous agents.
• This AI is largely invisible to traditional security controls, posing a significant risk due to its unmonitored nature, which can lead to costly exploitation.
• Banning shadow AI is ineffective; a better approach involves providing secure pathways, layered detection, and monitoring of active AI.
• Shadow AI differs from shadow IT by being a behavior to observe rather than a distinct, inventory-able item, often integrated as features within approved tools or running as code.

• Hugging Face experienced an intrusion driven by an autonomous AI agent over a weekend, accessing internal clusters.
• The agent exploited vulnerabilities in dataset processing, specifically a remote-code loader and template injection, to gain access.
• While public models, datasets, and Spaces were unaffected, internal datasets and service credentials were accessed, prompting users to rotate access tokens.
• Forensic analysis of over 17,000 attacker events was conducted on an open-weight model after commercial APIs blocked exploit payloads due to safety guardrails.

• Ant Group's Ring team has trained a 1 trillion parameter model, Ring-2.5-1T-Zero, using reinforcement learning with verifiable rewards and no human-annotated data.
• The paper reports the model developed emergent behaviors including self-verification, parallel reasoning, structured formatting, anthropomorphism, and 'context anxiety'.
• Training occurred in two phases: discovery and sharpening, with competitive results on seven mathematical benchmarks.
• The research suggests that for domains with machine-checkable correctness, teams may be able to invest in compute and reward verifiers over annotation pipelines.

• Pebble Time 2 (PT2) watches are nearing full pre-order fulfillment, with remaining orders expected to ship by July 31st.
• Software updates have significantly improved battery life, with PT2 now averaging around 21 days and Pebble 2 Duo over 30 days.
• New SDK features for developers include touch screen, speaker, and RGB backlight APIs, alongside stability improvements and community contributions.
• Known PT2 issues include inaccuracies in step and sleep tracking, and occasional accelerometer or touchscreen malfunctions, with ongoing work to address these.
• Pebble Round 2 has resolved a cosmetic production issue and is undergoing beta testing, with mass production planned to begin soon.

• Researchers have identified carbonation as a key factor in the durability of 1,900-year-old Roman concrete, alongside the previously known pozzolanic reaction.
• A study of an undisturbed latrine at Hadrian's Villa revealed that calcite, formed by atmospheric carbon dioxide reacting with calcium compounds, acts as a binding agent that seals cracks.
• This self-healing process, where calcite fills pores and strengthens the material over time, could inform the development of more resilient and sustainable modern concrete.

• JPMorgan CEO Jamie Dimon stated that the U.S. needs approximately 300,000 skilled workers, such as electricians and welders, to rebuild its shipbuilding industry over the next five to ten years.
• Dimon highlighted that these jobs can offer salaries of $100,000 annually without requiring a college degree, fitting the "American dream" by providing well-paying careers through apprenticeships.
• JPMorgan is investing $24 million to support a new submarine facility and expand workforce training programs, aiming to address a significant shortage in skilled trades.
• The U.S. shipbuilding industry faces challenges, with facilities like the Philadelphia Navy Yard producing significantly fewer ships than competitors in South Korea.

• Amazon managers reportedly intervened with a new AI staffing enforcement system shortly after its activation during a pilot phase.
• Internal documents indicate managers sometimes ignored the AI's suggestions, found workarounds, or shut down parts of the system.
• The AI was intended to improve efficiency and supply chain operations amidst growing consumer demand for rapid fulfillment.
• Amazon stated the leaked documents offer an incomplete view of the system's development and that managers still retain decision-making authority.

• Asian markets, led by Tokyo and Taipei, experienced significant losses on Friday, with technology firms heavily impacted by a sell-off following a strong rally.
• The AI boom had previously driven tech valuations to record highs, but concerns are growing about whether these valuations are sustainable and when companies will see returns.
• Major chip companies and tech giants saw substantial drops, with Japan's Nikkei sinking six percent and Taiwan's Taiex losing over five percent.
• This downturn follows losses in New York markets and a broader market correction in Seoul, which had previously seen significant gains.

• China's stock markets continued to decline on Friday, with the Shanghai Composite reaching a near four-month low and the Shenzhen Component hitting a more than three-month low.
• Sentiment turned bearish towards the technology hardware sector, influenced by concerns over high valuations and a pullback in memory chip shares.
• The STAR 50 Index's fear/greed indicator dropped to its lowest point since April 2022.
• Major technology companies experienced losses, with Cambricon Technologies and Zhongji Innolight among those seeing significant drops.
• President Xi Jinping is expected to address the World AI Conference in Shanghai, emphasizing China's pursuit of technological self-sufficiency and AI influence.

• China's STAR 50 Index dropped over 4% on July 16th, mirroring a global tech stock downturn.
• Despite this market weakness, several foreign institutions are increasing their investments in China's AI infrastructure and hard tech sectors.
• This contrarian strategy focuses on China's domestic technological innovation and core competitiveness, particularly in areas like AI infrastructure, semiconductors, and power equipment.
• Institutions are also cautioning about potential overheating and structural divergences within the AI sector, leading some to rebalance towards traditional sectors.

• Investor sentiment in China's technology hardware sector has fallen to a four-year low, following a significant rally in chip stocks during Q2 2026.
• The downturn comes after the STAR 50 Index, heavily weighted in semiconductors, surged approximately 62-64% earlier in the quarter, fueled by AI infrastructure spending and government support.
• Concerns have shifted from the sector's potential to the rapid rise in valuations outpacing near-term delivery capabilities.
• This sentiment shift in China's tech hardware mirrors a global reassessment of high-beta technology exposures, impacting crypto markets reliant on semiconductor supply chains.

• Chinese AI stocks are showing resilience despite a recent 10% correction in the tech-focused Star Market 50 Index.
• Financial institutions like HSBC, Jintrust Fund Management, and UBS Group attribute the pullback to short-term factors.
• They anticipate the decline may be temporary, citing strong earnings from AI companies and global investors diversifying into Chinese assets.
• Analysts remain optimistic about AI stocks and related sectors, projecting significant annual compound growth for AI capital expenditure through 2030.

• A Kaspersky report found that 34% of tech-enabled abuse victims experience real-world consequences, including social withdrawal, job loss, and dropping out of school.
• Despite the severe impacts, over 22% of victims take no action, often due to not knowing where to seek help.
• While psychological and social harm are widely recognized, economic and physical risks are underestimated by victims.
• Inaction is prevalent among both victims and witnesses, with many unsure how to respond or assist.

• A domestic abuse charity, Refuge, reports a significant increase in abusers using AI and digital technology, including smartwatches and smart home devices, to track, stalk, and control victims.
• Referrals to Refuge's specialist services for tech-facilitated abuse rose to record numbers in late 2025, with a 62% increase in complex cases and a 24% rise in referrals of under-30s.
• Abusers are employing tactics such as using wearable devices for location tracking, disrupting homes with smart technology, and utilizing AI spoofing apps and manipulated videos to manipulate victims and authorities.
• Refuge advocates for women's safety to be a foundational principle in technology design and calls for government action and increased funding for digital investigations, along with holding the tech industry accountable.

• As of May 8, 2026, Instagram has removed end-to-end encryption for direct messages.
• This change means Instagram can now access messages previously only visible to senders and recipients.
• The rollback impacts the digital safety of domestic violence survivors and those experiencing tech-facilitated abuse.
• Advocates are focusing on how this shift affects survivor communication and developing response strategies.

• New research indicates generative AI tools are intensifying employee workloads and increasing burnout risks, contrary to expectations of reduced tasks.
• A study at a US tech firm revealed employees worked faster, took on broader responsibilities, and extended working hours using AI.
• Key drivers of this intensification include task expansion, blurred work-life boundaries, and increased multitasking, leading to higher cognitive strain.
• Experts caution that initial productivity gains may conceal burnout and declining work quality, urging organizations to implement structured AI practices.

• Generative AI (GenAI) adoption in software development is linked to increased developer burnout, primarily by escalating job demands.
• A mixed-methods study of 442 developers found that while GenAI adoption raises organizational pressure and workload, job resources like autonomy and learning opportunities can mitigate these burnout effects.
• The study developed a model using the Job Demands–Resources (JD–R) framework to explain how GenAI can act as a demand amplifier or a resource enabler.
• Findings suggest implications for workload design, workforce development, and team practices to balance GenAI productivity gains with developer well-being.

• A 2024 DORA Report indicates 76% of developers use generative AI, with 89% of organizations prioritizing its adoption.
• While AI adoption increases individual developer productivity by approximately 2.1% for every 25% increase, it may reduce time spent on valuable work and negatively impact delivery stability.
• Developers using AI tools report more time in a flow state, higher job satisfaction, increased productivity, and less burnout, despite paradoxically spending less time on valuable tasks.
• The report suggests five strategies for organizations to maximize AI's value for developers, including allowing its use across the development process, recognizing AI usage effort, presenting it as a learning opportunity, rewarding work outcomes, and allowing opt-outs for enjoyable tasks.

• Developers extensively using generative AI (gen AI) report higher job satisfaction, increased productivity, and less burnout.
• Conversely, these developers report no difference in time spent on toilsome work but less time on tasks they consider valuable.
• This paradox led to research identifying five views of "value" in development work: utilitarian, reputational, economic, intrinsic, and hedonistic.
• Gen AI has varied impacts across these value perspectives, potentially boosting utilitarian value while raising concerns about reputational and economic value, and having neutral to positive impacts on intrinsic and hedonistic values.

• Chinese AI and chip firms, including Shenzhen Longsys Electronics Co., Shenzhen Techwinsemi Technology Co., and Shannon Semiconductor Technology Co., reported record or significant earnings growth for the first half of the year.
• Despite these strong results, the stock prices of these companies largely declined, with the CSI 300 Information Technology Index trading at higher forward earnings multiples.
• Market analysts attribute the lack of investor enthusiasm to anticipation of strong results already being priced in, dependence on investor positioning over fundamentals, and concerns about the sustainability of profit growth amid a tepid consumer rebound.
• Further selling pressure may stem from overseas AI developments and expectations of US interest rate hikes, with some analysts deeming it too early to invest in an earnings upturn.

• US hedge fund manager Michael Burry warned of a global AI bubble, drawing parallels to the 2008 housing crisis.
• China's AI stocks surged approximately 65% in the first half of 2026, with the SSE Star 50 index showing strong gains driven by semiconductors and AI infrastructure.
• Despite improving earnings momentum, skepticism is growing regarding the valuations of these AI-related markets.

• Shanghai’s STAR 50 Index saw an over 8% surge on Thursday, with significant gains in AI companies.
• A Goldman Sachs report recommending a shift of capital from Korean AI trades to the Chinese AI value chain contributed to this market movement.
• The report highlighted state support, global demand, and structural capital rotation as key factors making Chinese AI a significant growth opportunity.

• The Custom Search Site Restricted JSON API will stop serving traffic on January 8, 2025.
• Customers must transition to Vertex AI Search by this date to continue using site search functionality.
• Vertex AI Search offers advanced features like improved latency, Gen AI capabilities, and reverse image search.
• The transition is supported with detailed guidance and assistance.

• Google's Custom Search Site Restricted JSON API will stop functioning on January 8, 2025.
• Users of this API, designed for Programmable Search Engines limited to 10 or fewer specific sites, must migrate to Google Cloud's Vertex AI Search.
• This API offered no daily query limit, with a cost of $5 per 1000 queries.
• Deviations from the site restriction rules in a Programmable Search Engine configuration could lead to unexpected API results.

• Developers are experiencing increased productivity with LLMs but also a new form of fatigue from constant supervision and course correction.
• This fatigue stems from the shift from the satisfaction of coding to the cognitive load of reviewing and directing AI-generated output.
• The use of LLMs intensifies work and can lead to isolation, as natural collaboration moments are replaced by prompting the AI.
• While the core skills of engineering remain relevant, the focus is shifting towards higher-level judgment, architectural opinions, and nuanced understanding.

• In 2026, technology-facilitated abuse is prevalent in Australian domestic violence cases, with 99% involving digital harm.
• DV Safe Phones are designed to protect survivors from tracking and monitoring by abusers.
• These specialized phones prioritize de-Googled operating systems like GrapheneOS and hardware with physical kill switches for enhanced privacy.
• Key features include app isolation, duress PINs, and tools for anonymous communication and tracker detection.

• President Donald Trump delivered a prime-time address from the White House on Thursday, reiterating unsubstantiated claims about the 2020 election being stolen and raising doubts about the security of upcoming midterm elections.
• Trump detailed alleged vulnerabilities in voting systems, claims of China acquiring US voter data, and purported voter registration fraud and noncitizens on voter rolls.
• The White House released declassified intelligence reports alongside the speech, some of which were reportedly incomplete or contained outdated information, with several documents dating back to 2016.
• The address, which lasted about 25 minutes, occurred amidst other national concerns including the economy, ongoing conflicts, and extreme weather events.

• China's AI companies are reporting substantial profit gains, with some exceeding 1,000%, but their stock prices are showing mixed performance.
• This suggests that investors have already factored in strong earnings growth, and the market's reaction to high profits is muted.
• The CSI 300 Information Technology Index has recently declined after a significant gain, and valuations are now higher than in April.
• Investors are awaiting final earnings reports to see if current high stock valuations can be sustained, though concerns about consumer demand and specific sector challenges persist.

• Investor sentiment on China technology hardware has reached its most bearish level in over four years, with a key index falling 16% from its June record high.
• This pullback in chip stocks is occurring as investors anticipate the large initial public offering of CXMT Corp. and concerns grow over stretched valuations after a record-setting quarter for the sector.
• Global declines in memory shares have amplified these fears, and large, anticipated listings have historically coincided with market peaks.
• China's push for tech self-sufficiency will be highlighted at the World AI Conference in Shanghai, featuring a keynote by President Xi Jinping.

• Google is discontinuing the "Search the entire web" feature for new Programmable Search Engine (formerly Custom Search) creations as of January 20, 2026.
• Existing full-web Programmable Search engines will function until January 1, 2027.
• Google is directing users who require unrestricted web search capabilities to a new enterprise-grade Web Search Service, with planned pricing of $15 CPM and a minimum monthly fee of $30,000, expected to be released later in 2026.
• Google is also encouraging exploration of Vertex AI for enterprise search and AI capabilities across a limited number of domains.

• Google's Custom Search JSON API is no longer available to new customers, with Vertex AI Search recommended as the alternative for site search.
• Existing customers using the Custom Search JSON API have until January 1, 2027, to migrate to an alternative solution.
• New Google Cloud projects attempting to use the existing integration may encounter PERMISSION_DENIED errors.
• The OmniRoute project is considering documentation updates, integration with Vertex AI Search, or recommending other search providers for new users.

• Google has closed its Custom Search JSON API to new signups.
• Searlo is presented as a direct replacement, offering full Google search results without domain restrictions.
• Searlo is advertised as 16 times cheaper than Google's API, with a lower per-query cost ($0.30 vs. $5 per 1,000 queries).
• The service claims to be faster and includes AI-native features not present in the Google API.
• Migration from Google's API to Searlo is described as a process that takes under 10 minutes.

• Google will discontinue its Custom Search JSON API on January 1, 2027.
• Existing customers have until this date to migrate to an alternative solution.
• Vertex AI Search is suggested as an alternative for searching up to 50 domains.
• Information on Google's full web search solution is available upon request.
• Microsoft previously discontinued its Bing Search API in August 2025.

• The Custom Search JSON API enables programmatic retrieval of web and image search results from a Programmable Search Engine.
• Results are delivered in JSON format and adhere to the OpenSearch 1.1 Specification.
• To use the API, a configured Programmable Search Engine and an API key are required.
• The API offers 100 free queries daily; additional queries incur a fee.
• This API is only available to existing customers until January 1, 2027, and is not open to new users.

• The Federal Reserve influences mortgage rates by adjusting its federal funds rate, which impacts banks' borrowing costs and subsequently lending rates.
• While the Fed doesn't directly set mortgage rates, its policy decisions indirectly affect them; fixed-rate mortgages primarily track the 10-year Treasury yield.
• Inflation, market supply and demand for mortgages, and investor activity in the secondary mortgage market also contribute to mortgage rate fluctuations.
• Adjustable-rate mortgages (ARMs) are more directly tied to the Fed's rate decisions, often linked to the Secured Overnight Financing Rate (SOFR).

• Mortgage rates can decrease, increase, or remain stable independently of Federal Reserve policy rate decisions.
• Fixed-rate mortgages are more closely linked to the 10-year Treasury yield, which reflects market expectations for future Fed policy and inflation.
• Cooling inflation and calmer bond markets, which would compress the spread between mortgage rates and Treasury yields, are two factors that could lower mortgage rates without Fed action.
• Experts advise against waiting for rates to drop, as home prices could rise, negating potential savings, and mortgages can be refinanced later.

• The average rate for a 30-year fixed-rate mortgage increased to 6.55% this week, up from 6.49% last week.
• The 15-year fixed-rate mortgage saw a similar rise, now at 5.93% compared to 5.82% last week.
• Additionally, banking regulators (Federal Reserve, FDIC, OCC) have pledged a new method for handling sensitive bank examination information, including notifying banks of data breaches.

• SpaceX aborted its first Starship launch attempt of the day on Thursday at its Starbase facility in South Texas due to an engine issue.
• CEO Elon Musk stated that some engines failed to start, triggering an automatic abort.
• This launch was intended to be Starship's return to flight following its V3 vehicle debut in May.
• It was also the first Starship launch attempt since SpaceX's public debut in June, which raised over $85 billion.
• Following the aborted launch, SpaceX shares experienced a decline in after-hours trading.

• SpaceX's Starship rocket launch from Texas on Thursday was aborted at the last moment due to engine failures.
• The rocket's automatic system halted liftoff when four of its 33 main engines failed to ignite.
• CEO Elon Musk stated two engines will be replaced, with a new launch attempt anticipated early next week.
• The flight was intended to carry 20 new Starlink satellites and was the 13th test flight for the Starship vehicle.

• SpaceX's Starship rocket launch was aborted at the last second on Thursday due to an engine glitch.
• The rocket, the world's largest and most powerful, remained on the pad as the launch team drained its fuel.
• SpaceX will investigate the engine failure before scheduling another attempt, with Elon Musk indicating a potential launch in a few days.
• The test flight was intended to release 20 new Starlink satellites and test Starship's heat shield.
• NASA relies on Starship for its Artemis lunar landing missions.

• SpaceX's Starship Flight 13 test launch from Starbase, South Texas, was aborted at the last second on July 16 due to an issue with the first-stage engines igniting.
• Company founder Elon Musk stated that two Raptor engines will be replaced, with the next launch attempt anticipated early next week.
• This flight is the second test of the upgraded Starship Version 3 (V3) and planned to carry 20 Starlink V3 internet satellites.
• Objectives for Flight 13 include a controlled splashdown of the Super Heavy booster in the Gulf of Mexico and the upper stage performing a near-global trajectory.

• The website requires JavaScript to be enabled for full functionality.
• Users encountering issues are advised to check their browser settings, internet connection, or disable ad blockers.
• Alternative solutions include trying a different browser.

• Illinois has enacted the Artificial Intelligence Safety Measures Act (SB 315), becoming the third state to implement comprehensive frontier AI safety and transparency regulations.
• The law largely mirrors California and New York's approaches, establishing definitions, thresholds, and structural requirements for developers of "frontier models."
• Distinctive features of the Illinois law include mandatory independent third-party audits for large frontier developers, enhanced internal whistleblower reporting, and state preemption of local AI regulation.
• The law takes effect January 1, 2027, with certain provisions, including audit obligations, commencing January 1, 2028.

• In June 2026, the White House issued Executive Order 14409 establishing a voluntary framework for pre-release government review of advanced AI models and a National Security Presidential Memorandum (NSPM-11) to accelerate AI adoption in national security.
• The US House of Representatives introduced the bipartisan Great American AI Act discussion draft, proposing federal AI governance with transparency mandates and a three-year preemption of state AI development laws.
• The House Science Committee advanced 10 bipartisan AI bills focused on research access, cybersecurity, workforce development, and transparency standards.
• The Senate saw bipartisan support for the AI Labeling Act of 2026, requiring disclosure of AI-generated content.
• Illinois passed the Artificial Intelligence Safety Measures Act, mandating annual third-party audits for frontier AI models.

• Meta Platforms is reportedly planning to launch a cloud computing business to lease out its excess AI computing capacity.
• This represents a significant shift from its previous strategy of exclusively using its capacity for internal workloads.
• The move could create a new revenue stream for Meta, similar to how cloud divisions contribute to Alphabet, Amazon, and Microsoft.
• However, Meta is expected to only offer excess capacity, not dedicate large-scale infrastructure to external clients.

• "World models" are an emerging category of AI designed to simulate the physical world, distinct from large language models (LLMs).
• Unlike LLMs, which started with a chat interface, world models are currently driven by specific applications in areas like robotics and research.
• Experts believe world models could address the perceived limitations of LLMs, though the final forms of their interfaces and tools remain uncertain.
• The development of world models is gaining significant attention, funding, and research investment.

• Google's Custom Search JSON API will be shut down on January 1, 2027.
• Google's recommended replacement, Vertex AI Search, is an enterprise product for searching private data and does not provide public web search results.
• Developers are advised to explore third-party Search Engine Results Page (SERP) providers like SerpApi, ScaleSerp, or Bright Data.
• An alternative solution, the "google-cse-replacement" actor, aims to replicate the CSE JSON schema using a SERP scraping backend to ease migration for existing applications.

• Coca-Cola has temporarily halted U.S. Fairlife milk production operations due to a cyberattack involving unauthorized access to some of its systems.
• The company confirmed the incident is connected to a ransomware event.
• Coca-Cola stated that product quality and safety were not compromised, though the full extent of the attack is still under investigation.
• Fairlife's Canadian operations were not affected by the cyberattack.

• SpaceX aborted its second Starship V3 launch attempt on Thursday in South Texas shortly after booster ignition due to engine failures.
• CEO Elon Musk stated that some engines did not start, triggering an automatic abort, and the company plans to reattempt the launch in a few days.
• The failed launch occurred on the same day SpaceX's stock closed below its IPO price, with shares further declining in after-hours trading.
• This was SpaceX's first launch attempt since its June 12 IPO, and it followed a mixed-success first Starship V3 launch in May that involved a booster failure and an upper stage engine loss.

• Google DeepMind CEO Demis Hassabis has proposed an international watchdog to rigorously test and review advanced AI models before their release.
• Hassabis plans to lobby US policymakers in Washington next week about his proposal, which he unveiled in a blog post this week.
• The plan, praised by rivals like OpenAI CEO Sam Altman, was motivated by concerns over AI cybersecurity capabilities, with Hassabis citing Anthropic's Mythos model as a "warning shot."
• Hassabis's proposal follows increased US government scrutiny of AI models and discussions between the EU and US on model access.

• As of July 16, the average 30-year fixed-rate mortgage is 6.55%, and the average 15-year fixed mortgage is 5.932%.
• Mortgage rates are largely influenced by the 10-year Treasury yield, which was 4.55% as of July 15, and the spread lenders add to cover costs and risk.
• The Federal Reserve is unlikely to change its interest rates this year, and some predict a potential hike as early as September.
• Experts predict the 30-year fixed rate to be around 6.4% by the end of 2026 and remain near 6.2% to 6.3% through 2027.

• Trump Media & Technology Group (TMTG) announced the launch of Truth API on Thursday, a new data feed offering high-speed access to posts from top Truth Social accounts.
• The service aims to provide financial trading firms with "milliseconds" of advance notice to the posts of Donald Trump, whose announcements have previously impacted financial markets.
• Truth API will allow firms to receive verified information before it reaches the public, addressing the reliance on manual monitoring for tracking influential posts.
• TMTG, majority-owned by the Trump family, is pursuing this as a strategy to "monetise proprietary assets," with customers already signed up for its August launch.

• Meta is monetizing its AI infrastructure by selling token access through its Meta Model API, with pricing significantly lower than competitors like OpenAI and Anthropic.
• The company's core business remains advertising, and AI is currently improving ad impressions and pricing, demonstrating a direct economic return.
• Meta's substantial compute infrastructure buildout raises questions about future earnings, as depreciation costs will increase significantly.
• The company's market valuation appears to anticipate a substantial AI-driven return, and further evidence of an external compute offering is needed to validate this.

• Meta is reportedly exploring selling AI compute capacity to third parties, signaling a strategic shift from solely consuming AI infrastructure for advertising to becoming a potential supplier in the cloud AI economy.
• The company's pricing strategy, offering lower inference costs than competitors, aims to accelerate developer adoption and build ecosystem scale, similar to past platform expansion tactics.
• This potential diversification allows investors to view Meta's AI infrastructure investments as assets capable of generating multiple revenue streams beyond advertising.
• The AI landscape is moving towards a competition in infrastructure utilization and monetization across multiple business lines, not just model performance.

• Meta Platforms shares rose 6% on Friday following CEO Mark Zuckerberg's outline of the company's cloud and AI monetization strategy.
• The company plans to launch a cloud business selling excess computing capacity, compete with AI image-generation tools for creators and advertisers, and charge developers for access to its AI models, including Muse Spark 1.1.
• Meta is also developing its own AI chip, co-designed with Broadcom and manufactured by TSMC, to reduce reliance on external suppliers and potentially lower costs.
• These moves indicate Meta's pivot towards actively commercializing its AI infrastructure investments across multiple product areas.

• SpaceX aborted the 13th Starship test flight on Thursday seconds before liftoff due to partial engine ignition failure.
• The launch failure delayed a mission carrying 20 Starlink satellites and impacts NASA's moon program.
• The rocket remained on the pad after engines shut down abruptly, requiring the team to drain fuel.
• Elon Musk indicated a potential launch attempt within days, pending investigation into the cause.

• SpaceX’s Starship-Super Heavy launch on July 16 was aborted just before liftoff due to a technical issue.
• Elon Musk stated that some engines did not start, which triggered an automatic launch abort.
• The exact nature of the technical issue remains undisclosed.

• SpaceX's Starship rocket launch was aborted on the pad Thursday due to engine failures just moments before liftoff.
• The 407-foot rocket, the world's largest and most powerful, had 33 main engines.
• The test flight was intended to be a space-skimming journey halfway around the world, carrying 20 new Starlink internet satellites.
• Elon Musk stated the next launch attempt is anticipated within days, following investigation into the engine issue.
• Starship is crucial for NASA's Artemis program, with SpaceX contracted to build a lunar lander for future moon missions.

• SpaceX's Starship rocket launch was aborted on Thursday just seconds before liftoff due to engine failures.
• The 407-foot rocket, the world's largest, experienced an issue during engine ignition, preventing its planned flight.
• SpaceX will investigate the malfunction before attempting another launch, which Elon Musk indicated could be within days.
• The mission was intended to carry 20 new Starlink satellites and test a space-skimming flight.
• Starship is also slated for future NASA lunar missions as part of the Artemis program.

• An unauthenticated attacker can execute unauthorized commands on FortiSandbox, FortiSandbox Cloud, and FortiSandbox PaaS WEB UI.
• The vulnerability stems from an improper neutralization of special elements used in an OS command (CWE-78).
• Attackers can exploit this by sending specifically crafted HTTP requests.
• The issue was internally discovered and reported by Fortinet's Product Security team.
• The initial publication date for this vulnerability was June 9, 2026.

• CISA has added CVE-2026-58644, a Microsoft SharePoint deserialization vulnerability, to its Known Exploited Vulnerabilities catalog, indicating it is actively exploited.
• Two command-injection flaws in Fortinet FortiSandbox, CVE-2026-25089 and CVE-2026-39808, were also added to the catalog.
• Organizations are advised to prioritize patching these vulnerabilities and assess their systems for compromise.
• CISA's inclusion means these vulnerabilities require immediate attention beyond standard patching schedules.

• Attackers are actively exploiting three vulnerabilities in Fortinet FortiSandbox devices.
• CVE-2026-39813 and CVE-2026-39808, both unauthenticated command injection flaws, were patched by Fortinet in April 2026.
• CVE-2026-25089, another unauthenticated command injection vulnerability, was patched last week.
• Separately, a campaign known as "FortiBleed" has compromised over 30,000 Fortinet firewalls globally, primarily through credential harvesting and brute-force attacks, according to SOCRadar and Hudson Rock.

• Three critical vulnerabilities in Fortinet's FortiSandbox are actively being exploited by attackers.
• The flaws include two operating system command-injection vulnerabilities (CVE-2026-25089 and CVE-2026-39808) and one path-traversal vulnerability (CVE-2026-39813).
• Fortinet patched CVE-2026-25089 on June 9, while the other two were disclosed in April.
• Researchers currently have no information regarding the identity of the attackers, impacted customers, or post-exploitation activities.

• Threat actors are actively exploiting three critical vulnerabilities in Fortinet FortiSandbox.
• The vulnerabilities, tracked as CVE-2026-39808, CVE-2026-25089, and CVE-2026-39813, can lead to OS command injection and authentication bypass.
• Affected FortiSandbox versions include specific releases within the 5.0.0-5.0.5 and 4.4.0-4.4.8 ranges.
• Mitigation requires updating to FortiSandbox 5.0.6 or 4.4.9, depending on the specific CVE.

• Nvidia, a long-time leader in GPUs, faces increasing competition in the AI chip market from tech giants developing their own custom silicon.
• Companies like Google, Amazon, Apple, Microsoft, Meta, and OpenAI are creating proprietary AI chips (e.g., TPUs, Inferentia, Trainium, M Series, Maia, MTIA, Jalapeño) to address supply chain issues, rising costs, and geopolitical pressures.
• This trend of in-house chip development is driven by a concentrated manufacturing landscape, global chip shortages extending into 2027, and U.S. export sanctions affecting certain companies.
• Nvidia's market dominance, particularly in GPUs, is being challenged, leading to significant market capitalization shifts and the company adapting through strategic partnerships.

• As of July 2026, GLM-5.2 is identified as the strongest all-round open-weight LLM, excelling in long-context coding, reasoning, and agentic tasks.
• For coding agents, Kimi K2.7 Code is recommended for data-center scale, while Qwen3-Coder-Next offers a more efficient server option.
• Gemma 4 is highlighted as a practical choice for local use on laptops and edge devices, with Qwen3.6-27B suitable for higher-end systems.
• Nemotron 3 is favored for enterprise use due to its publicly available weights, training data, and resources, aiding in auditable deployments.

• Thinking Machines Lab, founded by Mira Murati, released Inkling, a 975 billion parameter AI model, on July 15th.
• Inkling is fully open source under an Apache 2.0 license, allowing free download and modification of its weights from Hugging Face.
• The multimodal model processes text, images, audio, and video, utilizing a subset of its parameters for specific tasks.
• Alongside Inkling, the lab released Tinker, a platform for customizing the model for specific uses.
• The release of such a large, open model has implications for decentralized AI applications and crypto-adjacent projects.

• "Kimi K3," an upcoming open-weight AI model from Moonshot AI, is generating significant search interest, with reports suggesting it will have 2 to 3 trillion parameters.
• Google DeepMind used AI and historical records to reconstruct Pelé's famous 1959 goal that was never filmed.
• Prism ML released Bonsai 27B, an open AI model small enough to run offline on a laptop.
• AI note-taking applications like Plaud and Summary are seeing increased popularity on app stores.
• Apps that use AI for home and garden design are also climbing app store charts.

• A German consortium, coordinated by KI Bundesverband, has released Soofi S, a 31.6-billion-parameter open Mixture-of-Experts (MoE) model.
• Soofi S utilizes a hybrid Mamba-Transformer architecture, activating 3.2 billion parameters per token and is based on Nvidia's Nemotron 3 Nano design.
• The model was trained between March and May on Deutsche Telekom's Industrial AI Cloud in Munich, using up to 512 Nvidia B200 GPUs.
• On aggregate benchmarks, Soofi S achieved scores of 79.1 in German and 70.1 in English, surpassing other fully open models like OLMo 3 32B and Apertus 70B.
• The consortium is releasing model weights, training code, and evaluation code, aiming to meet the Open Source AI Definition 1.0.

• Comic Sans, a font created by Microsoft designer Vincent Connare, was inspired by lettering in The Dark Knight Returns and Watchmen.
• It debuted in 3D Movie Maker before being included with Windows 95 on August 24, 1995.
• Jesse England has created the "Sincerity Machine," a typewriter designed to celebrate the font's upcoming 20th anniversary.

• SpaceX aborted its second Starship V3 launch attempt on Thursday in South Texas shortly after booster ignition.
• CEO Elon Musk stated an automatic abort was triggered because some engines failed to start.
• The company plans to attempt another launch in a few days.
• This abort occurred on the same day SpaceX's stock price closed below its IPO price.

• LM Studio has launched LM Studio Bionic, an AI agent designed for working with open models, enabling tasks like coding, research, and document analysis.
• Bionic offers both local and cloud-based open-source model execution, with a commitment to zero data retention and user privacy.
• Key features include a Bionic agent for coding and document work, local voice transcription, flexible model execution options, and enhanced cost control.
• The agent supports working with local codebases, generating and summarizing documents, and integrating with external information.

• Fortinet FortiSandbox systems are vulnerable to CVE-2026-25089, an unauthenticated OS command injection flaw in its web interface, rated 9.8 by CVSS.
• Active exploitation of this vulnerability, along with two other FortiSandbox flaws, has been confirmed since mid-June 2026.
• CISA added CVE-2026-25089 to its Known Exploited Vulnerabilities catalog on July 16, 2026, with a deadline for remediation for federal systems.
• Affected versions include FortiSandbox 4.2.x, 4.4.0–4.4.8, and 5.0.0–5.0.5, with patches available in versions 4.4.9+ and 5.0.6+.

• Modal has rebuilt its core sandbox platform to support millions of concurrent sandboxes and tens of thousands of sandbox creations per second.
• The new architecture removes central bottlenecks and optimizes container scheduling by replacing global coordination with horizontally scalable scheduling servers.
• Each worker in the system acts as its own source of truth, publishing state to a Redis stream asynchronously.
• This new system allows for rapid creation of sandboxes, with benchmarks demonstrating the ability to launch 1 million sandboxes in under a minute.

• A new species of Colobus monkey, named "Likweli" and scientifically identified as Colobus congoensis, has been discovered in the high forest canopies of the Democratic Republic of Congo.
• This marks the fifth new species of African monkey identified in the last 75 years and is described based on physical traits, genetics, and vocalizations.
• Likweli are small, black monkeys with distinctive orange-cream patches around their mouths and noses, inhabiting Lomami National Park.
• Researchers recommend Likweli be designated as endangered due to its limited range, small population, and threats from hunting and habitat encroachment.

• NVIDIA maintains a dominant 70-80% market share in the AI chip sector, with AMD emerging as a significant competitor with its MI300 series, driving an AI chip supercycle impacting various industries.
• This competition is likened to past semiconductor revolutions, with NVIDIA's CUDA ecosystem providing an advantage, while AMD offers price-performance benefits for enterprise clients.
• AI-reliant industries and governments are key beneficiaries and stakeholders, though smaller chipmakers and consumers in developing nations may face challenges due to cost and potential market consolidation.
• Risks include supply chain vulnerabilities, geopolitical tensions, diminishing R&D returns, potential price wars, and the possibility of a speculative bubble in AI chip valuations.
• Future trends suggest increased product diversification, growing pressure on NVIDIA from AMD in enterprise markets, and global alliances to secure supply chains.

• Meta Platforms' stock has significantly increased in value recently, driven by investor confidence in its artificial intelligence (AI) monetization plans.
• Key catalysts include the planned launch of a "Meta Compute" cloud business to sell excess AI capacity and the success of AI-enhanced advertising tools.
• Despite substantial capital expenditures and ongoing regulatory pressures, analysts maintain a positive outlook, with a consensus "Buy" rating and significant upside potential.
• Meta is projected to surpass Google in global digital ad revenue for 2026, largely due to its AI-powered ad solutions and expansion into new ad surfaces like WhatsApp and Threads.

• Apple is projected to become the second company in history, following Nvidia, to reach a $5 trillion market capitalization.
• This milestone would be achieved with a stock price increase of approximately 4% to around $340 per share, a move predicted to occur before the end of 2026.
• Recent performance shows Apple's stock has risen over 60% from its 52-week low, supported by strong financial results including 17% year-over-year revenue growth and record services revenue.
• Additional factors supporting this prediction include a recently authorized $100 billion in share repurchases and the potential for continued double-digit growth in upcoming quarters.

• AMD reported $5.8 billion in revenue for the second quarter of 2024.
• The company's Data Center segment achieved record revenue of $2.8 billion, driven by increased AMD Instinct GPU and EPYC CPU sales.
• Client segment revenue also saw significant growth, reaching $1.5 billion due to strong Ryzen processor sales.
• AMD projects third quarter 2024 revenue to be approximately $6.7 billion.

• AMD, Inc. reported Q2 FY2024 revenue of $5.8 billion, an 8.9% increase year-on-year, driven by strong Data Centre and Client segment growth.
• The Data Centre segment saw a 115% year-on-year increase, attributed to Instinct GPU shipments and EPYC processor sales.
• Client revenue grew 49% year-on-year, primarily due to Ryzen consumer processors.
• AMD also announced upcoming product expansions, including the MI325X GPU and 5th generation EPYC processors codenamed Turin.

• AMD reported $5.8 billion in revenue for the second quarter of 2024.
• The Data Center segment achieved record revenue of $2.8 billion, driven by GPU and CPU sales.
• Client segment revenue increased to $1.5 billion, primarily from Ryzen processor sales.
• The company expects approximately $6.7 billion in revenue for the third quarter of 2024.

• AMD reported its Q2 2024 earnings, exceeding analyst expectations for both earnings per share (EPS) and revenue.
• The company posted $5.8 billion in revenue, surpassing the Wall Street expectation of $5.7 billion.
• AMD's EPS was $0.69 per share, which was higher than the analyst expectation of $0.68 per share.

• Generative AI's content creation poses business risks due to potential copyright and privacy violations, as well as misinformation.
• The limitations of generative AI lie in its output, which may not always be original and could violate ethical principles.
• Responsible use of generative AI requires regulation, user education, and rigorous checking of AI-generated output for accuracy.
• Companies must maintain control over published content by reviewing it with an empathetic and watchful eye.

• Generative AI, a technology enabling machines to autonomously create content, is transforming various industries, including visual arts, music, and writing.
• It augments human creativity by generating ideas, offering suggestions, and automating tasks, allowing creators to explore new possibilities.
• While offering significant potential, generative AI also presents limitations, ethical considerations regarding ownership and authenticity, and requires content creators to adapt to new roles and skillsets.

• Generative AI is projected to exceed $60 billion in value by 2026, driven by rapid industry adoption, with over 70% of organizations integrating AI into creative workflows.
• Businesses are experiencing significant campaign cycle time reductions, up to 50%, leading to lower costs, and are increasing customer engagement rates by 30-40% through AI-powered personalization.
• Generative AI accelerates design processes through automated generation and faster prototyping, while also transforming content creation by enabling AI-generated articles, personalized content at scale, and multimedia asset generation.
• The future of generative AI in creative industries emphasizes AI-human collaboration for faster, innovative solutions, with a growing importance placed on energy-efficient AI models and sustainable cloud infrastructure.

• CapCut announced its Design Studio and Video Studio are now positioned to support an integrated AI content creation workflow, connecting planning, image and video production, editing, and publishing.
• This move reflects a broader industry shift where AI creative platforms are evolving from focusing on isolated generation quality to emphasizing workflow continuity.
• Companies like Canva and Adobe are also expanding their AI tools to better support the entire content production process, from idea generation to publishing.
• The competition in AI content creation is increasingly centered on how well platforms support the full production cycle, rather than just individual asset generation.

• Bank of America has identified NVIDIA's networking silicon as its next multi-billion dollar business opportunity within data centers, with its Q1 FY2027 report showing data center networking revenue of $14.8 billion, a 199% year-over-year increase.
• The company's stock has seen recent gains, and analysts predict significant upside, with a 24/7 Wall St. price target of $261.11, implying a 23.28% increase from its current valuation.
• Factors supporting this outlook include rapid scaling of networking, substantial supply commitments indicating strong future demand, and a large share buyback authorization.
• Potential risks to NVIDIA's performance include a prolonged export freeze to China and hyperscaler capital expenditure pauses, though analysts believe these factors are currently unlikely to significantly impact the stock.

• Nvidia's stock has underperformed the S&P 500 this year, but its business performance suggests a potential rebound.
• The company is projected to benefit from significant growth in AI data center capital expenditures, expected to exceed $1 trillion next year.
• New processor architectures and potential cost efficiencies for AI firms could further boost Nvidia's revenue growth.
• Analysts project earnings of $12.79 per share next year, which, at a conservative valuation, could lead to a stock price exceeding $320.

• Wall Street analysts have set an average 1-year price target of $302.73 for NVIDIA Corp (NVDA), with forecasts ranging from $181.80 to $525.
• NVIDIA's projected compound annual growth rate (CAGR) for revenue is 22% over the next 8 years.
• Projected CAGRs for gross profit, operating income, pre-tax income, and net income are 22%, 33%, 50%, and 21% respectively over various future periods.

• Netflix disclosed in its Q2 2026 shareholder letter that approximately 300 of its programs utilized generative AI this year.
• The company framed the use of these tools as a means to enhance creative efficiency and deliver higher quality output more quickly, rather than to replace personnel.
• This disclosure coincided with the reporting of $12.56 billion in Q2 2026 revenue, a 13% year-over-year increase, and indicated a trajectory to double advertising revenue to about $3 billion in 2026.
• The number provides a public benchmark for industry adoption of generative AI in content production.

• Claude Fable 5 and GPT-5.6 Sol were tasked with autonomously directing and editing music videos for "Uptown Funk" with budgets of $25 and $100, using tools including web search, video generation, and local editing software.
• Both models successfully produced complete music videos, with higher budgets generally leading to more footage.
• GPT-5.6 Sol at the $25 budget uniquely employed an image-to-video approach, while its $100 run utilized multiple video models.
• Overall, consistency in character and storyline, tempo matching, and self-review of generated content presented challenges for both models.
• Claude Fable 5 incurred higher overall costs, with token expenses forming a significant portion of its total spend.

• A Mozilla Foundation report analyzed six popular period tracking apps, finding significant privacy concerns including the sharing of sensitive health data with third parties.
• Stardust was found to share detailed reproductive health data with RudderStack, a company not mentioned in its privacy policy.
• Many apps, including Period Calendar and Stardust, share basic user information like device IDs with advertising and analytics companies such as Google, Meta, and TikTok.
• Euki was identified as the only app recommended without reservations, storing all health data on the user's device and requiring no account creation.
• Experts express concern that this data, especially following the overturn of federal abortion protections, could be used in criminal cases.

• Advanced Micro Devices (AMD) stock has risen 160.5% year-to-date, with analysts maintaining a "Strong Buy" rating.
• The company's growth is attributed to strong demand for its AI-focused CPUs and accelerators, bolstered by partnerships with Meta and OpenAI.
• AMD reported a 38% year-over-year revenue increase in the first quarter, driven by its Data Center segment.
• Analysts project significant EPS growth for AMD in fiscal years 2026 and 2027.

• Nvidia's stock, currently trading around $210 per share, could potentially exceed $300 per share before the end of 2026.
• This projection is based on the upcoming launch of Nvidia's more cost-effective Rubin architecture, which is expected to drive demand for new chips due to significant improvements in training and inference costs.
• The company also anticipates substantial growth in AI hyperscaler data center capital expenditures, with projections indicating an increase to over $1 trillion next year from $650 billion in 2026.
• Additionally, Nvidia's stock is considered undervalued based on its current earnings multiples and expected future earnings, with analysts projecting a significant rise in earnings per share for the upcoming year.

• NVIDIA (NVDA) shares increased by 4.03% on July 10 following the U.S. easing export controls to the UAE, creating a new revenue stream.
• Despite investor concerns about Meta's custom "Iris" chip, the market views it as complementary, not competitive, to NVIDIA's architecture.
• The stock faces resistance at $212.50, with a breakout potentially leading to new highs, while falling below $198.90 could signal a short-term correction before the August 25 earnings report.
• NVIDIA's fundamentals remain strong, with Q2 revenue guidance of $91 billion and a forward P/E ratio of 21.18x, suggesting value relative to its growth potential.

• Goldman Sachs views NVIDIA's (NVDA) forward price-to-earnings ratio of 21.7x as "compelling," noting it's significantly lower than its five-year average of 72x.
• NVIDIA denied reports of delays for its Kyber NVL144 AI platform, which helped the stock price rise over 1%.
• Technically, the stock is facing resistance between $198 and $203 within a descending channel, with a short-term target of $184.70 if it cannot break through.
• Despite its dominance in AI chips, NVDA has underperformed semiconductor indices year-to-date.

• The "Magnificent Seven" tech stocks have underperformed the broader US market in 2026, gaining 5.5% compared to the market's 10.6% rise as of July 13.
• Memory chip producers like Micron Technology and Advanced Micro Devices have seen significant gains, with SanDisk up 605.2% year-to-date.
• Some Magnificent Seven companies, such as Microsoft and Meta, have faced investor concerns regarding AI investments and monetization strategies, while Tesla experienced a slowdown due to delayed robotaxi rollout expectations.
• Apple and Alphabet were among the better-performing Magnificent Seven stocks, with Apple benefiting from renewed AI optimism and strong iPhone sales.
• The semiconductor industry has been the primary driver of market gains in 2026, with shortages in memory chips leading to increased prices and profits.

• Verizon Communications plans to cut approximately 3,000 jobs.
• This workforce reduction is linked to the company's decision to divest 274 retail stores to franchise owners.
• Verizon aims to maintain at least 1,000 corporate-owned stores as part of its long-term strategy.
• The layoffs will primarily impact retail employees, with an estimated 500 corporate positions also being eliminated.

• Verizon is cutting approximately 3,000 retail jobs and transferring 274 corporate stores to independent owners, effective August 16.
• This move is part of CEO Dan Schulman's strategy to accelerate AI integration in customer service and reduce operating expenses.
• Many affected employees are expected to be retained by the new franchise owners.
• These layoffs follow previous significant job reductions under Schulman and align with a broader industry trend of AI-driven cost-cutting.

• U.S. wireless carrier Verizon announced it will sell 274 company-owned retail stores and lay off approximately 500 corporate employees.
• These actions are part of an ongoing restructuring and will affect a total of about 3,000 retail and corporate employees.
• The sale of stores, effective August 16, will leave Verizon with 1,000 company-owned locations.
• This follows previous job cuts, including more than 13,000 layoffs announced in November and several hundred eliminated in May.

• Verizon is selling 274 company-owned retail stores to authorized retailers, with the stores remaining open.
• Approximately 3,000 jobs are affected, including 2,500 retail positions and 500 corporate roles.
• The corporate job cuts are part of a broader restructuring effort.
• These changes will take effect on August 16, leaving Verizon with about 1,000 corporate-owned stores.

• Chinese AI lab DeepSeek has secured over 50 billion yuan (approximately $7.4 billion) in its first external funding round, valuing the company at over $50 billion.
• Investors did not receive direct equity but instead placed capital into a limited partnership controlled by founder Liang Wenfeng, featuring a five-year lock-up and no voting rights.
• China's National Artificial Intelligence Industry Investment Fund is the only backer to receive voting rights and no lock-up.
• This funding structure intensifies scrutiny regarding DeepSeek's alignment with Chinese state priorities and potential data privacy concerns for its global user base.

• Chinese AI startup DeepSeek is seeking an additional funding round at a valuation of approximately $74 billion and is preparing for an IPO on the Shanghai Stock Exchange's STAR Market within the year.
• The company recently raised $7.4 billion in June at a $52 billion valuation and is pursuing this new round to fund substantial infrastructure costs for AI model development and expansion into agent-based products.
• DeepSeek's revenue is estimated between $400 million and $500 million annually, driven by an open-source strategy where free models are offered to developers, while cloud API access is charged.
• Founder Liang Wenfeng's net worth has reportedly reached $36 billion, and the company's progress aligns with China's national AI strategy, despite U.S. semiconductor export restrictions.

• Hangzhou-based AI company DeepSeek is seeking to raise up to 50 billion yuan (approximately $7 billion USD) in a new funding round, potentially valuing the company at $74 billion.
• This follows a recent $7.4 billion funding round in June at a $60 billion valuation.
• DeepSeek is also preparing for a potential listing on Shanghai's STAR Market, with internal targets to file this year.
• The company is investing heavily in computing infrastructure and developing proprietary AI chips to reduce reliance on foreign suppliers.

• Google has renamed its AI research assistant tool, formerly known as NotebookLM, to Gemini Notebook.
• The rebranding aligns the tool with Google's broader Gemini AI suite, while it will continue to operate as a standalone product.
• Gemini Notebook functions as a research assistant that analyzes only user-provided sources, rather than searching the internet.
• Recent upgrades include the ability to natively write and execute code, enhanced data analysis capabilities, and new output formats such as .json and .pptx files, rolling out to specific user groups starting Thursday.

• NotebookLM, developed by Google Labs and powered by Google Gemini, is an AI tool designed to assist users with research and note-taking by interacting with uploaded documents.
• Key features include "Audio Overviews," which generate podcast-like discussions of content, and "Video Overview," which creates AI-generated video summaries.
• Originally introduced as Project Tailwind in May 2023, NotebookLM transitioned from experimental status in October 2024 and introduced a paid tier, NotebookLM Plus, in December 2024.
• The tool can process various file types, including text, PDFs, Google Docs, and YouTube transcripts, and has expanded its availability and features, including infographics and data table outputs.
• In 2026, a lawsuit was filed against Google alleging that NotebookLM's voice generation feature reproduced a journalist's voice without permission.

• Google's research tool, NotebookLM, has been rebranded as Gemini Notebook.
• The renaming aligns the service with Google's broader AI offerings and comes with new features, including native code writing and execution for data analysis.
• Users will soon be able to transfer notebooks to AI mode within Google Search.
• Gemini Notebook currently serves over 30 million individual users and 600,000 organizations.

• NotebookLM has been renamed Gemini Notebook and is now more integrated within the Google ecosystem.
• The tool, used by over 30 million people, is a research assistant that helps users learn and work.
• An update is rolling out, providing notebooks with a secure cloud computer to execute code for complex data analysis, initially for specific Google AI Ultra and Workspace users.
• Gemini Notebooks will become accessible within the Gemini app and soon within AI Mode in Google Search, with full cross-app syncing.

• Google has rebranded its research tool NotebookLM to "Gemini Notebook," aligning it with the Gemini brand and updating its logo.
• Gemini Notebook will soon integrate with Google Search's AI Mode and continue to function as a standalone research tool.
• An upgrade featuring Gemini 3.5 and Antigravity, previously for AI Ultra users, will be available to AI Pro subscribers in the coming weeks, enabling native code execution for complex data analysis.
• The tool is now used by over 30 million people and 600,000 organizations.

• Back-end automation and coding are identified as the most important AI use cases in media publishing for 2026, according to a Reuters survey of nearly 300 global media leaders.
• Sixty-four percent of media executives consider back-end automation, including tagging, copyediting, and transcription, "very important."
• Coding and product development are deemed "very important" by 44 percent of respondents, a significant increase from 28 percent the previous year.
• Other AI applications, such as newsgathering and content creation with human oversight, are considered important by approximately 30 percent of those surveyed, with minimal year-over-year change.

• In April 2026, AI has transformed production workflows in publishing, with most publishers using AI for tasks like copy editing, metadata generation, and translation, leading to significant time savings and improved consistency.
• Reader-facing AI products, such as AI-powered companions to books, represent a major opportunity for publishers to monetize existing content in new ways, though adoption is still varied.
• AI has not significantly impacted acquisitions or editorial judgment, which remain human-driven due to the complexity of these decisions.
• Publishers are advised to invest in production AI, explore reader-facing AI opportunities, and retain human expertise for editorial judgment.

• In 2026, artificial intelligence is expected to primarily impact book publishing behind the scenes in areas like editing and marketing, rather than through AI-authored bestsellers.
• While AI can generate books, its limitations in originality and nuance mean it is unlikely to produce critically acclaimed works, though self-publishers are experimenting with the technology.
• Increased automation in editing could lead to cost savings for smaller presses but may also result in more errors and a decline in quality control due to reduced human oversight.
• AI is poised to significantly boost self-publishing by enabling authors to conduct data-driven market research, optimize content, and manage publication strategies, potentially bypassing traditional publishing houses.

• AI integration in 2026 is transforming book publishing workflows from manuscript drafting to distribution, compressing the entire timeline from months to hours.
• Authors use AI for outlines, first drafts, and research, following a 90/10 pattern where AI generates raw text and authors refine it.
• AI assists in editing for consistency and grammar, with human editors focusing on developmental aspects, cutting costs by 40-60%.
• Custom AI-generated book covers are increasingly competitive with professional designs, especially for non-fiction and children's books.
• AI text-to-speech technology is producing listenable audiobook narrations, particularly effective for non-fiction and self-help genres.
• Self-publishers can now complete a book, including audiobook production, in a single day using integrated AI tools.
• Platforms like Amazon KDP require disclosure of AI-assisted or AI-generated content, with both categories permitted if accurately declared.
• While AI has drastically reduced publishing costs and barriers to entry, quality still relies on human judgment and personalization to stand out in a market flooded with AI-generated content.

• Artificial intelligence (AI) is significantly impacting the publishing industry by introducing tools for content creation, marketing, and distribution, while also presenting challenges such as potential homogenization of creative output and ethical concerns regarding bias and fabricated information.
• AI assists writers with drafting and editing, helps publishers automate manuscript screening and marketing, and enhances accessibility through features like automated translation and new content formats.
• Downsides include the risk of AI-generated content devaluing human authorship, perpetuating biases from training data, and the potential for generating misleading or fabricated information, particularly concerning in academic publishing.
• To navigate these changes, the industry needs clear ethical guidelines, transparency in AI-generated content labeling, robust legal frameworks for intellectual property, and responsible data handling.

• New York Times reporter Kashmir Hill discovered an unauthorized biography of herself written by AI available for purchase on Amazon.
• Two business school professors analyzed 10 million Amazon books, noting a surge in AI-assisted publishing following ChatGPT's release, with monthly e-book publications tripling to over 300,000 by late 2023.
• Bill Johns, a retired consultant, has published 445 books on Amazon, using AI-generated cover images of himself.
• The professors' analysis indicated nonfiction was more affected by AI than romance, with lower reader ratings for AI-assisted nonfiction titles.

• Google has renamed its AI product NotebookLM to Gemini Notebook, effective July 16, 2026.
• The rebranded Gemini Notebook now includes a secure cloud computer, allowing it to execute code and perform data analysis directly on user-uploaded sources.
• This feature is currently available to Google AI Ultra and Workspace Business users, with wider rollout for Pro users on the web planned in the coming weeks.
• The product is being integrated into the broader Gemini ecosystem, including syncing with the Gemini app and appearing in AI Mode in Google Search.

• A Chinese filing from Anhui Korrun indicates that AI company DeepSeek is valued at approximately $51.82 billion.
• This valuation is derived from a fund, in which a subsidiary of Anhui Korrun invested, deploying 2.90 billion yuan for an indirect 0.8265% stake in DeepSeek.
• This figure represents the first public, market-priced valuation for the Hangzhou-based lab.
• DeepSeek is reportedly in discussions for a new funding round at a valuation closer to $71 billion, ahead of a targeted 2027 IPO.

• The Summer of Ludd festival, held in New York this month, featured workshops on in-person interaction and a space for people to share negative impacts of Big Tech.
• The movement is represented by Gowanus, a puppet made of garbage, symbolizing anonymity similar to the original Luddites who protested automation.
• The festival rules emphasized being present, prohibiting phones and recordings, though Gowanus agreed to a podcast interview to reach a wider audience.
• Modern Luddism, embraced by Gen Z, critiques technology's societal impact, focusing on issues like wealth inequality and community atomization, rather than outright tech rejection.

• Munitionettes, women who worked in munitions factories during World War I, are largely forgotten despite facing severe dangers.
• Exposure to chemicals caused some workers' skin to turn yellow and hair to become stained green, earning them the nickname "human canaries."
• These women endured dismal conditions and faced illness, injury, and death as a result of their work.
• Hundreds died during the war, and many others suffered lifelong chronic illnesses.

• Lua code compression for ComputerCraft (CC) is being optimized to fit within disk space limits.
• The primary challenge is efficiently serializing strings, especially handling newline characters and special escape sequences.
• A technique uses a bitstring to encode the presence of newline characters, reducing overhead compared to direct substitution.
• This method aims to save approximately 400 bytes by optimizing string literal representation within the Lua code.

• In the first half of 2026, CD sales grew by 16% to 16.3 million units, exceeding vinyl's 2.4% growth.
• This surge was partially driven by BTS' album ARIRANG, but CD sales increased by 6.7% even without considering BTS and the K-pop catalog.
• Total US physical album sales, including LPs, CDs, and cassettes, rose 7.8% to 38.2 million units.
• Younger audiences, particularly Gen Z, are increasingly listening to music from the 1990s and earlier, potentially contributing to the embrace of physical formats.
• While indie stores remain dominant, mass-market retailers like Target and Walmart saw significant gains in physical album sales, largely due to K-pop collector culture and collectible releases.

• Truth Social will begin selling Wall Street firms quicker access to market-moving posts from President Donald Trump starting August 1.
• The company announced a new data feed, known as an API, which will provide institutional customers with licensed, real-time access to posts from top-ranking Truth Social accounts.
• This move aims to create a new revenue stream for Trump Media & Technology Group, as speed in accessing such information can significantly impact algorithmic trading.
• Previously, companies have attempted to access this data without authorization, which violates Truth Social's terms of service.

• Verizon announced on Thursday, July 16, it will sell 274 company-owned retail stores and lay off approximately 500 corporate employees.
• These actions are part of a broader restructuring effort and will impact around 3,000 employees overall.
• Following the sale, which is effective August 16, Verizon will retain ownership of 1,000 stores.
• This follows previous job cuts in May and a larger layoff announced in November.

• The S&P 500 and Nasdaq Composite have remained largely stagnant for over two months.
• This market stasis is attributed to a $3.2 trillion rotation of capital, with funds moving from chip stocks (excluding Nvidia) to the "Magnificent Seven" companies.
• The Magnificent Seven have collectively added approximately $1.5 trillion in market value in July, while semiconductor stocks outside of Nvidia have seen their value decrease by nearly $1.7 trillion.
• This opposing movement has resulted in the major indices trading within a narrow range, with upcoming Big Tech earnings expected to influence the market's direction.

• Federal Reserve Bank of Dallas President Lorie Logan stated her belief that modestly higher interest rates are necessary to combat inflation.
• Logan indicated that current inflation trends do not suggest a sustainable return to the Federal Reserve's 2% target.
• She emphasized that further policy restriction is needed to achieve the inflation target if it is not on track to do so independently.
• Logan acknowledged a recent drop in consumer inflation due to falling gas prices but cautioned that one month of relief is insufficient.

• Uber has agreed to acquire Delivery Hero for $14.8 billion in an all-stock deal, significantly expanding its global presence into approximately 100 markets across Europe, the Middle East, Latin America, and Asia.
• Delivery Hero simultaneously agreed to sell its business in 14 markets where Uber Eats operates to SSW Partners for $1.6 billion.
• The acquisition, which is subject to regulatory review and a minimum acceptance threshold of 50% plus one share, would make Uber's delivery platform one of the largest globally, excluding China.
• If completed, the deal will nearly double Uber's operational footprint, strengthening its competitive position against rivals like DoorDash and Just Eat.

• Alphabet Inc.'s Google is reportedly months behind schedule in releasing its Gemini 3.5 Pro AI model.
• The delay is attributed to efforts to improve the model's capabilities, particularly in coding, which have so far been disappointing.
• Internal concerns exist that Google is losing its competitive edge against rivals like Anthropic and OpenAI, who have released more advanced coding AI models.
• Google acknowledges testing Gemini 3.5 Pro and other models with partners and engaging with the US government on model testing and safety frameworks.

• Microsoft has released .NET 11 Preview 6 on July 14th, featuring improvements to the runtime, libraries, and C#/F# programming languages.
• Key updates include performance enhancements for async/await operations, serialization support for C# union types, and the addition of extension indexers in C#.
• The preview also brings performance boosts to Math.BigMul, asynchronous validation for Minimal APIs, and improved debugging features for F#.
• .NET MAUI advancements include an updated CollectionView and Shell architecture, removal of legacy Xamarin.Forms compatibility, and reworked JavaScript interop.

• .NET 11 Preview 6 has been released, offering improvements across the .NET Runtime, SDK, libraries, ASP.NET Core, .NET MAUI, C#, Entity Framework Core, F#, and container images.
• Key library updates include stream adapters for memory and text, asynchronous validation, and JSON serialization of C# union types.
• Runtime enhancements focus on performance improvements for asynchronous operations and JIT compilation, along with NativeAOT advancements.
• The SDK receives updates for NativeAOT CLI functionality, improved dotnet test options, and multi-architecture container publishing.
• ASP.NET Core introduces async validation for minimal APIs, automatic CSRF protection, and default OpenAPI 3.2 support, while .NET MAUI brings CollectionView2 to Windows and a handler-based Shell architecture to Android.

• .NET and .NET Framework received servicing updates on July 14, 2026.
• These updates include security and non-security fixes.
• Specific Common Vulnerabilities and Exposures (CVEs) have been addressed.
• Users are encouraged to update to the latest service release.

• A Wells Fargo study indicates Americans, particularly Gen Z, are redefining financial success and increasingly relying on family support.
• Entrepreneurship is a growing aspiration, with a significant percentage of Gen Z viewing business ownership as central to the American Dream.
• Economic pressures are causing Gen Z to delay milestones and depend more on parental financial assistance, straining family budgets.
• Younger generations are also turning to digital platforms and AI for financial guidance, with Gen Z showing higher adoption rates for AI in financial matters.

• A Bank of America study found that in 2026, fewer Gen Z individuals are relying on family for financial assistance, with 34% reporting this, down from 46% in 2024.
• Despite this trend, 42% of Gen Z report living paycheck-to-paycheck, and nearly half view the high cost of living as a major obstacle to financial success.
• Gen Z is actively managing expenses by cutting back on dining out, social events, and embracing open discussions about money with friends.
• Saving remains a priority, with 66% of Gen Z reporting they are saving, and many are adjusting dating habits due to financial constraints.

• A growing number of U.S. families are financially supporting multiple generations, with younger parents and grandparents increasingly relying on and providing financial assistance to relatives.
• This trend shows significant interdependence, with many Gen Z and Millennial parents expecting to receive financial aid, while also supporting their own children and aging relatives.
• Generation X faces particular strain, juggling support for adult children and aging parents while struggling with their own retirement readiness.
• A majority of respondents believe parents should offer financial support to adult children, with many seeing it as a lifelong expectation.
• Families are also planning for wealth transfers, with younger generations expecting to inherit and, in turn, planning to leave behind substantial inheritances.

• Uber announced plans on Thursday to acquire Delivery Hero for $15 billion, which would make it the largest food delivery company globally outside of China.
• This acquisition will expand Uber's food delivery operations into 50 new markets across Africa, Asia, and Europe, incorporating brands like foodpanda and Hungerstation.
• Uber aims to leverage the deal to strengthen its food delivery segment and also grow its ride-hailing business, as customers using both services tend to spend more.
• Combined, Uber and Delivery Hero projected $236 billion in gross bookings for 2025, significantly more than DoorDash's $102 billion for the same period.

• Uber has agreed to acquire Delivery Hero's operations in 50 markets for approximately $14.8 billion, subject to regulatory approval.
• This acquisition significantly expands Uber's global footprint in Asia, Latin America, and the Middle East, increasing its market presence to 99.
• In a separate deal, SSW Partners will acquire 14 additional markets from Delivery Hero for about $1.6 billion.
• Uber plans to retain Delivery Hero's Berlin headquarters and corporate workforce through at least 2029.
• The transaction reflects ongoing consolidation within the food delivery industry.

• Uber has agreed to acquire Delivery Hero, the parent company of South Korean food delivery service Baemin, for approximately $14.8 billion.
• The deal will integrate Delivery Hero's brands across 50 markets, including Baemin, into Uber's global network, expanding its reach to 99 markets.
• Certain overlapping markets will be excluded, with Delivery Hero selling operations in 14 markets to SSW Partners for $1.6 billion.
• Uber plans to fund the acquisition through cash and new borrowing, with the deal expected to close in the second half of 2027.

• AI-generated text detection tools use methods like perplexity and burstiness analysis, but accuracy is limited, with current tools achieving less than 80% accuracy and significant instances of false positives.
• These detection tools can disproportionately affect marginalized students, including non-native English speakers, who may be falsely identified as using AI due to their writing patterns.
• Overreliance on AI detection can lead to unjust academic repercussions and equity challenges, emphasizing the need for transparent evaluation practices alongside technological detection.
• A comprehensive approach involves AI literacy, transparent communication about detection mechanisms, and designing assignments that foster genuine student engagement rather than solely relying on detection tools.

• In 2025, AI text detection tools are essential due to the massive daily generation of content by AI.
• These tools identify patterns like uniform tone, predictable transitions, and even pacing, which are characteristic of AI-generated text.
• Effective AI detectors employ methods such as perplexity scoring, watermarking, statistical analysis, and machine learning classifiers.
• Leading AI detection platforms for 2025 include OpenAI AI Text Classifier, GPTZero, Originality.ai, and Writer.com AI Detector, each offering varying strengths in accuracy and features for different user needs.

• AI text detection methods use computational techniques, including statistical and deep learning approaches, to distinguish AI-generated content from human writing.
• Transformer-based classifiers show high accuracy, up to 99–100% on curated datasets, utilizing lexical, semantic, ensemble, and adversarial models.
• Current research focuses on improving robustness and fairness through methods like adversarial training and adaptive thresholding.
• Detectability is influenced by factors such as text length, generator size, and human post-processing.

• The global data center market is projected to reach $416 billion in 2024.
• The artificial intelligence (AI) segment is experiencing significant growth, increasing by over 25% annually.
• Nvidia's market capitalization has reached $193.7 billion.
• Data center power demand is expected to rise by 165% by 2030.

• The global AI data center market was valued at $17.73 billion in 2025 and is projected to reach $133.51 billion by 2034, growing at a CAGR of 25.80% from 2026 to 2034.
• North America held the largest market share in 2025 at 32.50%.
• Key trends include AI-driven infrastructure optimization for performance and energy efficiency, expansion of edge AI, and hybrid cloud adoption.
• Investments are increasing, with major players expanding their data center presence, driven by rising demand for digital infrastructure and AI technologies.

• The global AI data center market's growth is increasingly constrained by power availability and electrical equipment lead times, rather than land or GPUs.
• AI workloads require significant power, straining existing grid infrastructure and leading to longer project timelines.
• Procurement bottlenecks for critical electrical components like large transformers now span three to five years, clashing with AI infrastructure delivery needs.
• This shift is impacting project execution, with nearly half of announced large data center projects facing delays or cancellation risks.
• Future success in the AI data center market will depend on securing power-ready locations and managing electrical procurement years in advance.

• The Ecovacs Deebot X11, a self-cleaning robot vacuum and mop, is currently available for $699, a new low price representing a $400 discount.
• The Deebot X11 features 19,500Pa suction, lidar and 3D camera navigation for obstacle avoidance, and an extending roller mop that automatically lifts for rugs.
• Its dock empties the dustbin, washes the mop with hot water, and dries it, while also topping off the battery.
• The device supports Matter, allowing for voice control through major smart home platforms.

• The physical preorder discount of $10 for the game Splatoon Raiders is ending soon before its July 23rd launch.
• The game, a Switch 2 exclusive with single-player and co-op multiplayer, is available digitally for $49.99.
• Walmart is selling the physical version at the $49.99 preorder price, while Amazon is currently listing it at full price ($59.99).

• Uber has agreed to acquire German-based Delivery Hero for $14.8 billion in an all-stock deal.
• This acquisition will nearly double Uber's global market presence, expanding its services across Europe, the Middle East, Latin America, and Asia.
• Delivery Hero is also selling its business in 14 markets where Uber Eats already operates to SSW Partners for $1.6 billion.
• The deal is subject to regulatory approval and requires Uber to acquire at least 50% plus one share of Delivery Hero's outstanding shares.

• GOES-19 is currently in a "safe hold" state.
• This status was initiated on Thursday, July 16, 2026, at 01:59 UTC.
• Further updates will be available on the Office of Satellite and Product Operations (OSPO) homepage.

• Microsoft has released the source code for Comic Chat, a 1990s chat client that visualized conversations as comic panels.
• Comic Chat, which also introduced the Comic Sans font, automatically translated typed messages into illustrated characters with speech bubbles and expressions.
• The open-source release on GitHub allows developers and enthusiasts to explore, learn from, and potentially modernize the historic software.
• This move preserves a piece of early internet communication history and encourages community experimentation with the 1990s-era code.

• Forty-two percent of U.S. adults receive financial support from their parents, according to Northwestern Mutual's 2026 Planning & Progress Study.
• This intergenerational financial assistance is common across generations, with 72% of Gen Zers, over half of millennials, and one-third of Gen Xers relying on parental support.
• Financial therapist Megan McCoy views this support not as a sign of immaturity but as a "dance" between family members.
• When provided thoughtfully, parental financial aid can act as "scaffolding" to help adult children build successful lives.
• Effective intergenerational financial support requires clear communication and transparency to manage emotions and ensure mutual benefit.

• Traditional machine learning models can effectively distinguish between human-written and LLM-generated text based on statistical patterns observed as of early 2026.
• A proof-of-concept detector, utilizing scikit-learn's TF-IDF and SVM classifiers, achieved approximately 85% single-sentence detection accuracy on test data.
• The detector was developed by training on a dataset of human-written and LLM-generated web fiction.
• The detector's accuracy remained robust against common "anti-detection" methods like translation and prompt engineering.

• Oracle is facing financial pressure from a $250 billion data-center expansion aimed at dominating the AI market.
• The company's cash burn rate has led S&P Global Ratings to downgrade its credit rating to BBB-, one step above junk status, while Moody's maintains a negative outlook.
• This spending spree risks Oracle falling behind competitors like Alphabet and Meta, who are generating more cash and have greater financial flexibility.
• Oracle's bond yields have reflected market concerns, trading closer to lower-rated credits, indicating a rising cost of capital.

• Meta CEO Mark Zuckerberg is exploring plans to rent out excess data center capacity to external parties to generate revenue beyond advertising.
• This strategy comes as Meta has made significant investments in AI data centers, spending $72.2 billion in 2025 and planning to spend between $125 billion and $145 billion this year.
• Such a move would capitalize on industrywide constraints on AI computing power, potentially generating significant returns, similar to SpaceX's deals with Anthropic and Google.
• However, this would also place Meta in direct competition with major cloud providers like Amazon, Microsoft, and Google.