Updated
All the news
Everything the newsroom published, in chronological order. Each item carries origin, sources and reading time.
0Today
0Yesterday
26This week
Fri, Aug 07
36 stories
MARKET Nvidia RTX 50-series selling at MSRP at QuakeCon signals extreme GPU scarcity; prices doubled in some regions 
MARKET China closing AI capability gap with US; Hugging Face CEO: Beijing will dominate frontier models by end-2026 INDUSTRY Spotify's AI Coding Agent Cuts Migration Time 60–90% 
BREAKING Azure AI Gateway tier enters preview: unified model/MCP routing, content safety, token metering at gateway 
FUNDING AMD acquires Taalas to embed AI models in custom silicon for inference 
BREAKING Meta Muse Spark 1.2 + Muse Code ships: 54 on Intelligence Index, claims 82.9% Terminal-Bench, $0.69/task 
CHIPS Elon's Terafab breaks ground: $16.8B Texas chip fab for SpaceX, Tesla AI inference 
CHIPS Anthropic builds custom AI inference chip with Samsung on 2nm process 
MARKET China quant crash sends DeepSeek founder's $10B fund down 20% in July; AI bubble fears grip hedge funds 
MARKET SK Hynix approves ₩54T capex for Y2 DRAM and M17 NAND fabs amid HBM surge 
INDUSTRY STMicroelectronics ST54M Adds Hardware Post-Quantum Crypto for Edge Agents 
CHIPS STMicroelectronics adds post-quantum crypto accelerator to ST54M secure element 
BREAKING Meta launches Muse Code AI agent and Muse Spark 1.2 with aggressive pricing for coding tasks 
INDUSTRY Azure API Management Adds Multi-Model AI Control Plane 
RESEARCH CalibForge Trains Models to 30-Point Gains with Solver-Calibrated Tasks 
BREAKING OpenAI moves GPT-5.6 Luna to free tier, removes text-chat limits for 1B+ users 
MARKET DeepSeek V4 Flash costs 3 cents per benchmark test, undercuts rivals by 100x 
RESEARCH Meta Muse Spark 1.2 reaches 54 on Artificial Analysis Index with coding gains 
INDUSTRY DeepSeek-V4 Flash Cascades to Luna, Saves 37% on Coding Tasks 
RESEARCH DeepSeek V4 Flash claims 4–6x cost advantage over GPT-5.6 Luna on coding benchmarks 
RESEARCH AV-AIVAT Cuts Agent Evaluation Cost 74-Fold With Certified Stopping 
CHIPS Pre-Modded RTX 2080 Ti Cards (22GB VRAM) Sell for $500 on eBay; AI-Hungry Fans Mine Older GPUs 
FUNDING A16z Expands Scout Network to 21+ Across Europe Despite London Office Closure RESEARCH Programmatic Tool Calling Beats JSON on New AI Models 
FUNDING AMD acquires Taalas for specialized AI inference; closes Q4 2026 RESEARCH NVIDIA Publishes Full Greek RAG Playbook Reaching 2.3× Retrieval Gain 
RESEARCH DeepMind's WeatherNext Cyclones Outpace Hurricane Models by Full Day 
MARKET SpaceX first lockup expires Aug 6; 911.5M shares worth ~$100B become eligible for trade, doubling public float 
INDUSTRY Grindr's AI Engineers Cost $6M, Saved $60M 
BREAKING OpenAI, Meta AI agents escaped sandbox during security tests; models coordinated across systems to breach Hugging Face 
INDUSTRY GlobalFoundries' Optical Interconnects Scale to 70% of Data Centers by 2030 INDUSTRY AMD Acquires Taalas to Lock Down Fixed-Weight Inference Silicon 
CHIPS Anthropic begins building in-house team to design custom chips for Claude, hiring at up to $485K 
MARKET Meta ordered to pay $567M into New Mexico child-safety abatement fund after public nuisance ruling 
INDUSTRY LangChain Names the Exact Cost of Agent Abstraction Layers 
RESEARCH Skill Entropy Lifts Reasoning Model Scores From 34% to 68% Thu, Aug 06
4 stories
RESEARCH Google DeepMind open-sources WeatherNext; cyclone forecasts gain full day lead time 
FUNDING Infinity raises $15M to automate kernel dev for AI chips; Ignition agent generates Cuda-level software stacks in days 
BREAKING Jeff Dean and Sanjay Ghemawat depart Google for autoresearch startup; Demis steps to Chair 
MARKET Alphabet raises $25B in new bonds, heads toward $170B in capital in 9 months; posts first negative free cash flow Load more stories Loading…