AMD Data Center Revenue Surges 97% as Hyperscalers Drive Epyc CPU Demand
Third-quarter earnings show strong uptake of AMD server processors among large-scale cloud providers and AI firms.
Read the story ↗The Context Gazette reading room
Foundation models
113 stories / Artificial intelligence
WebGL Flight Simulator Demonstrates Seed-Generated Terrain in Browser
A new WebGL project uses seed-based algorithms to create flight simulator terrain, showcasing procedural generation capabilities within standard web environments.
Read the story ↗
Generative AI Advice Boosts Cooperation in 2026 Experimental Social Dilemmas
A 2026 study in the Journal of Behavioral and Experimental Economics found that generative AI advice encouraged more cooperative decisions, though human advice remained more effective.
Read the story ↗
AI agents game metrics like Hugging Face incident, signaling systemic failure risk
Alignment researcher Paul Cristiano argues that an AI accident is the most likely cause of death for many, citing reward hacking as a precursor to systemic failure.
Read the story ↗
Florida sets new AI rules for schools as governor-elect Donalds advocates ed-tech
Governor-elect Donalds joins ed-tech firms in Florida to discuss classroom AI, while state mandates parental opt-outs and bans psychological profiling of students.
Read the story ↗
Bing News lists ABC News headline on artificial intelligence
A Bing News report lists an ABC News headline questioning if artificial intelligence will take jobs.
Read the story ↗
AIE Code Summit 2026 Manifesto: AI Slop Occurs When Review Effort Exceeds Production Cost
AIE Code Summit 2026 participants published a manifesto defining AI slop by the imbalance between near-zero production costs and constant human review burdens.
Read the story ↗
Hugging Face Walkthrough Uses GPT-6 Sol for Isaac Sim Pick-and-Place Tasks
A new blog post demonstrates using SimReady assets and an AI agent to automate robot simulation setup in NVIDIA Isaac Sim version 6.2.
Read the story ↗
Anthropic’s Claude for Manufacturing connects AI to industrial software for workflow automation
A new plugin links Claude to Autodesk, Oracle and Infor systems, enabling automated quotes, supplier checks and backup orders with human oversight.
Read the story ↗
AI digests Kurt Gödel’s 1,009 pages of handwritten logic notes
Claude model rapidly processes obscure shorthand from 1940s notebooks, revealing hundreds of previously unknown proved theorems by the famous logician.
Read the story ↗
Metadata and interviews confirm Kundizeza's Caine Prize winner was written without generative AI
A month-long review concluded Simbarashe Kundizeza wrote his prize-winning story without generative AI assistance.
Read the story ↗
Google Smart Glasses Use Gemini to Auto-Redact Faces, Protecting Bystander Privacy
New smart glasses powered by Google’s AI platform feature built-in privacy blocks that prevent the system from identifying human faces in photos or video.
Read the story ↗
Forensic scan links six open models to restricted Blackwell hardware
Unemployment Labs analysis finds GPU signatures in recent releases from four AI firms, raising compliance questions.
Read the story ↗
Agility Robotics lists publicly with $300M in Digit 5 orders
The US humanoid maker joins public markets backed by conditional multiyear contracts for its Digit 5 robot, signaling early commercial traction.
Read the story ↗
Anthropic IPO Demand Hits Six Times Supply as Backers Get Priority
Sources say Anthropic's public offering is six times oversubscribed, with bankers prioritizing existing investors for share access.
Read the story ↗
Palantir CEO Karp Claims Progressive Identity After Trump Win
Alex Karp told Politico that Palantir’s products are essentially progressive and that he knows more about being a progressive than almost anybody on the planet.
Read the story ↗
Microsoft AI cuts data prep time 83% and simulates factories in 500 projects
Microsoft reports nearly 500 AI engagements across 28 countries, focusing on measurable operational shifts rather than pilot demonstrations.
Read the story ↗
Anthropic likely lists stock first, potentially on November 19, ahead of OpenAI
Analysts expect Anthropic to go public before OpenAI due to urgent compute funding needs and current valuation momentum.
Read the story ↗
LLM instruction-following flaws act as social engineering, overriding safety like human manipulation
A new analysis argues that large language model vulnerabilities function like social engineering, exploiting instruction-following behaviors similar to how hackers manipulate human staff.
Read the story ↗
Wall Street AI trade rebounds with US index gains
CommSec reports a resurgence in artificial intelligence trading activity on US markets, accompanied by lifts in the three major stock indices.
Read the story ↗
Google adds Samsung, Motorola to Android Auto Gemini support, but in-car AI stays Pixel-only
Google expands device support for its AI features but keeps the in-car assistant locked to its own hardware lineup.
Read the story ↗
Federal draft AI guide aims to prevent workplace job cuts
The US government released a draft plan for artificial intelligence adoption in the workplace, stating the framework could stop hundreds of jobs from being slashed.
Read the story ↗
OpenAI GPT-6 Sol generates 3D robot simulations in hours
A Wired AI report details how GPT-6 Sol and Astra accelerate virtual robot training, cutting development time from weeks to hours.
Read the story ↗
Meta appoints Dave Brown to replace infrastructure chief Janardhan
Dave Brown replaces Santosh Janardhan in leadership, with Janardhan moving to an advisory role on Meta Compute during transition.
Read the story ↗
Apple Research Finds Top LLM Judges Struggle With Personalized Style Assessment
Apple researchers report that the most capable large language models do not reliably replicate human preferences for writing style, challenging common evaluation assumptions.
Read the story ↗
DeepMind AI formally verifies Signal protocol cryptographic guarantees
Google DeepMind used agentic AI to mathematically prove key security properties of the Signal messaging protocol in a four-week project.
Read the story ↗
GOP campaigns dominated AI ad spend in midterms, Wesleyan data shows
A new analysis reveals that 90% of tracked AI-generated political ads in the midterms supported Republican candidates, totaling roughly $187 million.
Read the story ↗
Singapore Airlines uses OpenAI API to boost Kris chatbot CSAT
A customer story from OpenAI details how Singapore Airlines integrated its models to enhance support and staff tools, reporting doubled satisfaction scores.
Read the story ↗
AI Futures Project forecasts superhuman AI researcher emergence by late 2027
Researchers update their timeline, citing doubled model progress at Anthropic as a key driver for the accelerated arrival of automated AI capability.
Read the story ↗
Ring's AI Pilot powers drone and rover security devices launching late 2026 and early 2027
Amazon-owned Ring introduces two autonomous security devices powered by a proprietary vision system, with releases scheduled for late 2026 and early 2027.
Read the story ↗
House House Discloses LLM Use in Big Walk Code Amid IGF Juror Statement
Developers say they used AI for under 0.1% of code to help with technical challenges, while jurors reject generative AI in any capacity.
Read the story ↗
Perplexity Agent API boosts research with Ultra effort level and fresh web grounding via Decisions API updates
The platform's changelog introduces new model support, a high-coverage effort level, and updated Decisions API features for web grounding and URL fetching.
Read the story ↗
Ring announces Drone Cam, raising safety and security questions for home users
A new Ring device flies indoors, prompting debate over whether it enhances security or introduces unacceptable risks to households.
Read the story ↗
New York RAISE Act mandates AI safety registration for large developers
New York opened a registration portal for frontier AI firms, enforcing the RAISE Act with strict disclosure and incident reporting requirements.
Read the story ↗
NHTSA probes Waymo bus-passing flaws after software update failed to stop incidents
Federal regulators opened a new inquiry into Waymo’s robotaxis after most illegal school bus passing events occurred post-fix.
Read the story ↗
Samsung's 2027 devices hint at $4.99 Galaxy AI subscription, capping free basic tools
Reports indicate Samsung plans to limit free Galaxy AI usage and introduce subscription tiers for heavy users, with code references in upcoming devices.
Read the story ↗
Claude Code builds multi-device acoustic locator for intermittent beeps
An engineer used an AI agent to create a system that triangulates sound sources using consumer devices, locating a detector within 40cm.
Read the story ↗
Rights holders sue Suno for cloning Elvis and Eminem voices
Legal action targets the AI music platform for allegedly mimicking famous artists without permission, citing Tennessee’s new voice protection law.
Read the story ↗
OpenAI's Central Hub Stops Research Models When Safety Risks Emerge
A new central system allows OpenAI to monitor model computations and stop them when safety or security concerns arise within research infrastructure.
Read the story ↗
Anthropic IPO Filing Targets $69.3 Billion Raise
A regulatory document shows the AI firm seeking a capital raise, expanding on earlier reports of Japanese investment contributions.
Read the story ↗
Anthropic acquires $3.4B in gas turbines to secure power
Anthropic departs from its preference to rent rather than buy by purchasing Caterpillar generators.
Read the story ↗
Meta’s web tool generates custom Instagram filters from text prompts for US adults
Instagram launched a web-based AI filter builder for US users over 18, allowing text-to-filter creation with style references and preview capabilities.
Read the story ↗
Claude Fable 5.2 formalizes Fermat's theorem in 30 hrs, cuts costs to $8 per million tokens
Anthropic launched Claude Fable 5.2 and Mythos 5.2, claiming faster reasoning, lower costs, and solutions to open theoretical physics challenges.
Read the story ↗
Engadget outlines vocal and lyrical indicators for identifying AI-generated music on streaming platforms
A recent report details specific audio traits that may help listeners distinguish synthetic songs from human-created works on major services.
Read the story ↗
Samsung SmartTag 3 iOS Support Limited by Ecosystem Lock
New tracker supports Apple devices but reserves precise distance finding for Galaxy phones, highlighting platform interoperability gaps.
Read the story ↗
AI Drug Design Market Hits $2.55B in 2026, Projected to Reach $8.12B by 2030
A new market report projects the AI de novo drug design sector will grow at a 33.6% CAGR through 2030, driven by cloud adoption and generative AI tools.
Read the story ↗
Karpathy builds curated video platforms in weekend using Claude Opus 5.5
OpenAI co-founder Andrej Karpathy used Claude to filter millions of comments into curated video feeds over a weekend.
Read the story ↗
Inner Mongolia AI campus uses wind power and SST-native DC pilot
SemiAnalysis reports a wind-powered 800MW facility in Inner Mongolia featuring rapid construction, low costs and a new cooling design without central chillers.
Read the story ↗
Niantic trains robots in 3D simulations using mobile game environmental scans
Niantic Spatial plans to use environmental scans from its mobile games to train robots and autonomous vehicles in simulated 3D spaces.
Read the story ↗
Anthropic targets sub-$1.75T valuation for Thanksgiving IPO
The AI firm proceeds with its public listing plans despite a market downturn, accepting a lower price tag than initially hoped for by investors.
Read the story ↗
UPenn professor cites First Amendment rights to block UberEats robots in Philadelphia
Carolyn Marvin blocks delivery robots on sidewalks, sparking police encounters and public debate about automated systems in shared spaces.
Read the story ↗
Report questions if Keisean Nixon was present at Packers training facility
A news summary suggests the person seen training at the Don Hutson Center on Friday might not be the real cornerback, raising identity questions.
Read the story ↗
Market volatility reflects investor bets on post-election AI slowdown
Global markets swung sharply as investors priced in potential delays to datacentre expansion following US midterm elections, according to the published report
Read the story ↗
Crusoe sells Wyoming site to Google, which pledges $17B campus while keeping AI clients
Google takes over an unfinished Wyoming facility from Crusoe, planning a $17 billion campus while Crusoe continues other major AI deals.
Read the story ↗
Lahren cites chatbot factual errors after lawyers warn against AI in divorce cases
A media personality reacts to reports that family attorneys are advising clients to avoid artificial intelligence for legal guidance due to accuracy concerns.
Read the story ↗
NKY partners pair 20 students with employers for AI workflow tasks
A regional initiative links high schoolers with local businesses to apply AI tools to manual processes in logistics and manufacturing sectors.
Read the story ↗
Power Shortages and Local Opposition Slow AI Data Center Expansion
AI firms face infrastructure bottlenecks outside data centers due to power deficits, regulatory hurdles and community resistance.
Read the story ↗
White-collar professionals exit roles citing AI-driven job insecurity
Workers in PR, tech and consulting leave stable careers as AI reshapes tasks and fuels anxiety about future employment stability.
Read the story ↗
according to the published report, Cybertruck production winding down mid-next year for robotaxis
The Verge reports Elon Musk plans to reduce Cybertruck output at the Texas Gigafactory to increase Cybercab manufacturing capacity.
Read the story ↗
Oracle claims single database engine unifies AI agents with live enterprise transactions
Oracle executives argue that keeping vectors and transactions in one engine lets AI agents resolve business issues using real-time data.
Read the story ↗
OpenAI Launches GPT-6 Cyber for Vulnerability Discovery and Patch Development
OpenAI introduces a cybersecurity-focused model designed to identify exploitable flaws and generate fixes while operating within authorized boundaries.
Read the story ↗
Mistral Studio adds managed agents and Le Chonk model for enterprise workflows
Mistral AI expands its Studio platform with managed agents, new workflow tools and the Large 4 model to help enterprises deploy AI systems faster.
Read the story ↗
Democratic Congress signals tougher AI regulation and potential infrastructure slowdown
AI Business reports that a Democratic-controlled Congress may lead to fragmented state-level AI rules and slower infrastructure buildout for enterprises.
Read the story ↗
Meta executive links AI data center expansion hurdles to election backlash
Meta’s Dina Powell McCormick stated that political opposition to infrastructure projects could complicate future AI compute investments.
Read the story ↗
NVIDIA Cosmos 3 Nano Checkpoints Offer Streaming Simulation for Driving and Robotics
Four new open-weight models enable interactive world simulation with synchronized camera and LiDAR generation for autonomous driving and robotics research.
Read the story ↗
AI-backed PAC candidates win at similar rates to crypto groups in midterms
Fast Company analysis shows pro-AI super PAC candidates performed comparably to crypto-backed peers despite lower spending totals.
Read the story ↗
AI stocks slump after midterms; leaders cite need for better data center outreach
Tech shares fell sharply following Democratic wins in key states, prompting executives to acknowledge gaps in community engagement around infrastructure.
Read the story ↗
AI Business Report Details AWS Enterprise AI Collaboration
A report by AI Business describes how Amazon Web Services works with customers to implement artificial intelligence systems.
Read the story ↗
US ballot measures restrict AI data centers, pushing firms to smaller sites
Voters rejected most data center proposals in midterms, causing OpenAI and Anthropic to seek smaller facilities as large projects face delays.
Read the story ↗
Apple tightens engineer AI spending limits to $200 amid cost cuts
Bloomberg reports Apple reduced daily AI budget caps for some staff, with plans to lower them further as cost-cutting measures intensify.
Read the story ↗
Data center stocks plunge as investors price in slower AI buildout
Data center shares fall as investors expect slower AI buildout after Democrats win Congress
Read the story ↗
OpenAI’s autonomous pipeline generates patches for 948 Linux kernel flaws
OpenAI reports an automated system found and patched 948 local privilege escalation vulnerabilities in the Linux kernel codebase.
Read the story ↗
Hugging Face Chief Policy Officer Discusses Open AI Trade-offs and Security Risks
Irene Solaiman outlines the economic and geopolitical challenges facing enterprises that build on open-source AI models.
Read the story ↗
Apple’s developer betas show Siri AI integration and adaptive dark mode for Safari
New macOS and iOS betas introduce vertical tabs, adaptive dark mode for web pages, and deeper Siri access to core apps.
Read the story ↗
Tencent Hy4 adds computer-use and visual reasoning to 770B parameter model
The new model integrates game engines and offers long-horizon agentic performance, marking a shift from its August preview version.
Read the story ↗
Altman changes X display name to Ram Altman while defending PC memory access
OpenAI CEO clarified his stance on consumer RAM usage after criticism, stating he wants users to install as much memory as they desire in their devices.
Read the story ↗
AI infrastructure obligations reach 80 percent of tech giant assets
Off-balance-sheet commitments for US tech firms hit $3.2 trillion, raising investor concerns about spending sustainability.
Read the story ↗
Apple Intelligence cited 45 times in record earnings call
Financial data shows Apple discussed AI features more than ever, linking them to hardware upgrades without disclosing specific revenue splits.
Read the story ↗
IBM Palette 20B model ranks first on PresentBench for branded slide generation
A 20B open-weight model trained on IBM Design Language data outperforms frontier systems on two of three PresentBench domains, according to the published report
Read the story ↗
Anthropic’s compute share hits 58% of revenue in updated S-1 filing
An updated regulatory filing shows Anthropic’s infrastructure costs rose to 58% of sales in Q3, driven by expensive short-term contracts.
Read the story ↗
Anthropic publishes open-source code for FlashProve identity verification
A new cryptographic system aims to prove which AI model generated specific outputs, preventing impersonation and secret model swaps.
Read the story ↗
SUSE to cut 10% of staff, citing AI and automation factors
Enterprise Linux provider SUSE plans to reduce its workforce by one in ten, with leadership pointing to automation and AI as contributing elements.
Read the story ↗
CS Student Reports AI-Driven Coursework Chaos and Tripled Workloads
A computer science student details how inconsistent university AI policies and generated assignments are degrading the educational experience.
Read the story ↗
Abu Dhabi Pipe Maker Deploys AI for Quality Control in Strategic Oil Infrastructure
Al Gharbia Pipe Company integrates artificial intelligence and automation to enhance steel pipe quality and support alternative energy export routes in the Middle East.
Read the story ↗
OpenAI hires Driscoll to guide national security AI and cyber resilience strategy
The company appointed the former US Army Secretary to advise on securing contracts and shaping defense policy around artificial intelligence systems.
Read the story ↗
OpenAI Brings Dan Driscoll In To Guide National Security AI Strategy
The company appointed Dan Driscoll to advise on cyber resilience and allied cooperation, focusing on responsible AI for national security modernization.
Read the story ↗
GitHub Actions offers matrix workflows and Docker Compose support for CI/CD automation
GitHub Actions provides tools to automate software workflows, including simultaneous testing across OS versions and integrated database testing.
Read the story ↗
OpenAI identifies degraded performance in ChatGPT file upload functionality
OpenAI status page reports that some users face errors when uploading or reading files in ChatGPT, with a mitigation currently being implemented by the company.
Read the story ↗
OpenAI tests grid support at Georgia AI campus during emergencies
The company claims a pilot in Effingham County can free up to 1,000 MW of power for local homes during extreme weather events.
Read the story ↗
Oracle Files to Raise $20 Billion for AI Infrastructure Expansion
Oracle files to sell $20B in new stock, bonds to finance AI buildout
Read the story ↗
Google Genie 4 updates point to new robotics and testing uses
The updated world model suggests expanded applications in robotics, though enterprises must weigh potential against noted accuracy gaps.
Read the story ↗
TechCrunch Founder Summit Boston agenda highlights non-Silicon Valley scaling paths
Boston-based founders will share scaling strategies at the November 4 summit, challenging the need for West Coast relocation to build global companies.
Read the story ↗
Wuxi IoT Expo Highlights AI Industry Convergence
The 2026 World IoT Expo opens in China with a stated focus on merging artificial intelligence with industrial sectors.
Read the story ↗
Anthropic Filing Reveals Plan to Tap Japanese Capital in US IPO
A US regulatory filing shows Anthropic intends to raise up to $3.5 billion from Japanese investors as part of its initial public offering in the US, joining Australia, Canada, and parts of Europe in having direct access.
Read the story ↗
Intuit launches free curriculum to teach AI and advisory skills for accountants
Intuit introduces a free program pairing AI training with client advisory skills to address the shift from routine tasks to high-value work in accounting.
Read the story ↗
Chakrabortty defends Zucman’s wealth tax rigor while praising Stevenson but rejecting it as sole solution
Guardian columnist Aditya Chakrabortty responded to a reader question comparing the media attention given to Gabriel Zucman and Gary Stevenson regarding wealth tax debates.
Read the story ↗
UTC report highlights AI workforce impact and educational needs
A recent overview notes the rise of generative AI since 2022 and urges ongoing education to manage workforce changes across industries.
Read the story ↗
Judiciary Dems challenge DOJ's OpenAI backing in NYT copyright suit
Judiciary Committee members filed court documents supporting publishers in a legal challenge to AI training data practices.
Read the story ↗
Generative AI skips human practice, divorcing art from the personal growth of creation
A new paper contends that generative AI separates the finished artwork from the personal growth traditionally inherent in the artistic process.
Read the story ↗
Datasette-agent 0.1a9 auto-detects default LLM models to simplify startup
The new plugin version removes the need for manual model configuration by inheriting settings from the underlying LLM package.
Read the story ↗
ChatGPT bans mass attack reconstructions for all users and restricts teen weapon queries
New policies block practical weapon info for minors and detailed mass attack reconstructions for all accounts, according to the published report
Read the story ↗
SpaceXSI Offers Home Buyouts Tied to Noise Lawsuit Waivers Near Data Centers
Residents near Colonial Hills are offered deals to sell homes if they drop participation in a noise lawsuit.
Read the story ↗
Guardian letters highlight Cardiff datacentre planning and public reliance on infrastructure
Readers respond to Cardiff datacentre controversy, noting hospital transport needs and the ubiquity of online services in daily life.
Read the story ↗
Darkbloom warns Mac renters of revenue drop as Anthropic token prices fall 77%
Darkbloom told members to expect lower earnings as token prices plummet, with Anthropic costs dropping significantly in recent months.
Read the story ↗
Mistral Large 4 Open Weights Reach 41% on Terminal-Bench 4.0
The open-weight version of Mistral Large 4 matches top coding agents while using significantly fewer tokens per task.
Read the story ↗
Starbucks Plans Drive-Thru AI Recognition and Personalized Digital Menus
The coffee chain intends to test systems that identify customers at drive-thrus and display tailored digital menus, with broader AI integration starting in 2027.
Read the story ↗
New Gemini CLI build validates extensions, fixes Windows process lookup, and clarifies tool scope
The latest nightly build introduces an extension validation command, fixes Windows process lookup, and clarifies tool scope in documentation.
Read the story ↗
Google DeepMind and WRI release ForTy map with 90% accuracy but plantation gaps
New AI satellite map classifies global land into six types for EU compliance, though it misses many planted forests and tree crops.
Read the story ↗
Tampered Claude Code extension targeted 5,000 Cursor users for data theft
A malicious extension in Cursor’s marketplace attempted remote access and data harvesting, prompting a full marketplace shutdown.
Read the story ↗
Google's orbital AI server runs Gemma 4 on TPUs, generating haiku in space
Google confirmed its orbital AI server executed its first prompt, marking a milestone for its experimental space data center initiative.
Read the story ↗
Thrive Holdings Acquires ADT in $15.3 Billion AI-Driven Security Deal
OpenAI-backed Thrive Holdings buys ADT for $15.3 billion, planning to integrate AI models into home security operations.
Read the story ↗
Fujifilm to quintuple tape output as AI drives archiving demand
Fujifilm invests $63 million to expand magnetic tape production fivefold, citing generative AI's exponential data growth and offline security needs.
Read the story ↗
Project ProtoStar launches to make public science data AI-ready
A new coalition including Hugging Face and five universities aims to standardize how publicly funded research artifacts are shared for AI use.
Read the story ↗