AI news / automated archive.
Anthropic-Physical Intelligence acquisition rumors ignite AI Twitter, OpenAI's pre-release models accidentally breach Hugging Face, and Google ships three new Gemini models but skips 3.5 Pro. Plus: US threatens sanctions on Chinese AI, Jack Dorsey's Buzz challenges Slack, and Meta tests AI bedtime stories.
China's Moonshot AI unveils Kimi K3 to rival OpenAI and Anthropic, Beijing's AI companion law shuts down Doubao and Qwen, and Meta's Muse Spark 1.1 narrows the gap. Plus: AI agent monitoring challenges, Nvidia's $100M voice AI bet, and Anthropic's Pentagon tensions.
China's open-weights AI strategy dominates discussion as open models gain ground on proprietary alternatives. Moonshot AI unveils Kimi Work, Cursor explores agent swarm economics, Qwen drops Image 3.0, and US tech giants' AI debt hits $1.65 trillion amid the infrastructure spending race.
The head of the US AI safety agency resigns after just three months. Meta and Anthropic discuss a landmark $10 billion compute deal. Alphabet's Google is developing a new 'Frozen' chip for radically more efficient AI inference — and more in today's roundup.
Bezos-backed CuspAI raises £334M for AI materials discovery, China's Moonshot AI halts new Kimi subscriptions amid surging demand ahead of IPO, and US public health agencies begin testing OpenAI and Anthropic models — plus NVIDIA's $20T vision, Europe's anti-AI-slop platform, and TSMC's blockbuster AI-driven results.
Google's Gemini launch hits internal roadblocks, Apple Intelligence secures China approval with Alibaba and Baidu, and Databricks reaches a staggering $188 billion valuation — plus Chrome's stealth 4GB AI install, Current AI's open infrastructure push, Bunkerhill's agentic healthcare raise, and Kimi 3 closing in on Opus 4.8.
