Is Anthropic’s AI Development Speeding Up? What METR Found
METR estimates Anthropic improved about twice as fast on well-specified AI tasks, with substantial uncertainty. Here is what the review does and does not show.
METR estimates Anthropic improved about twice as fast on well-specified AI tasks, with substantial uncertainty. Here is what the review does and does not show.
After Texas elected Gina Hinojosa, proposed data center protections could affect Musk’s AI expansion. Reuters reports the risk; no statewide pause has been enacted.
OpenAI says Images 2.6 adds short GIF creation, more consistent image edits, and faster generation on average. Here is what is rolling out and what remains to be tested.
Google plans new K-12 AI controls, but schools that reject AI may also lose Google Search on managed student accounts. Here is what is planned.
Garry Tan says his AI agents cost up to $200,000 annually and can read personal notes. Here is what he reported, how YC’s Stray Light differs, and the privacy tradeoffs.
Goldman Sachs reportedly limits Claude Code and GitHub AI coding tools to engineers, with a strats exception. Here is what the policy does and does not say.
Marvell announced a $115B enterprise-value deal for Lumentum. See why optical links matter for AI data centers, and which deal numbers are projections.
Google ran a Gemma 4 model aboard a prototype satellite. What the haiku test proves, and what orbital AI infrastructure still has to solve.
Mistral released the weights for its trillion-parameter Large 4 model. Learn what open weights mean and how to read its cybersecurity claims.
New York has signed rules for AI companion chats with minors. See what features they restrict, the exceptions, and when the law takes effect.