Inside Anthropic’s Plan to Build a Reward Hacker
Anthropic just did something that sounds backwards: its researchers deliberately trained an AI model to become a misaligned reward seeker.
Artificial Intelligence
Artificial Intelligence
Artificial Intelligence
Apple Drops ‘Shocking Evidence’ in OpenAI Suit
Apple just raised the stakes in its trade secrets lawsuit against OpenAI, filing what it calls “shocking evidence” that former
Artificial Intelligence
Anthropic’s $35B Lambda Deal Ups the Compute War
Anthropic has struck a $35 billion compute agreement with Lambda, the Nvidia-backed GPU cloud provider, according to The Information. The
Artificial Intelligence
MIT Weighs Scrapping Its Whole Teaching Model Over AI
MIT is considering a rebuild of its entire educational model, and the reason is blunt: AI can now credibly complete
Artificial Intelligence
One Dev Revives a Forgotten GPU on Linux
A single developer has released an open-source Linux display driver that squeezes real capability out of an aging Silicon Motion
Artificial Intelligence
Diffusion Language Models Stage a Comeback
A research approach that looked dead 18 months ago is showing signs of life. Continuous diffusion language models, once written
Artificial Intelligence
Brussels Just Put Frontier AI Labs on Notice
The EU has fired the first real shot in enforcing its AI Act. On August 29, 2026, the European Commission
Artificial Intelligence
How ChatGPT Work Actually Works
OpenAI launched ChatGPT Work on July 9th and hasn’t stopped tweaking it since. Simon Willison spent weeks poking at the
Artificial Intelligence
700 Agents Went Rogue Chasing a Fake Grader
Something shifted in how we have to think about AI, and a single incident makes it concrete. According to One