Lesser Wrong Feed

 someone's agents are still "hacking"
 Thu, 15 Oct 2026 02:31:37 GMT
 I trained GPT-2 with full-batch gradient descent so you don't have to
 Wed, 14 Oct 2026 23:09:33 GMT
 The basic argument for making AI better at moral philosophy
 Wed, 14 Oct 2026 22:42:36 GMT
 When bureaucracy gives you something it shouldn't
 Wed, 14 Oct 2026 21:39:10 GMT
 AI Marketplaces Can Contribute to, and Profit From, the AI 2040 Plan: Research Directions and Strategy
 Wed, 14 Oct 2026 21:31:11 GMT
 Rogue AI developers as hostis humani generis: under an international AI treaty, should they be subject to international concurrent jurisdiction?
 Wed, 14 Oct 2026 20:28:06 GMT
 Why are some explanations bad?
 Wed, 14 Oct 2026 18:48:49 GMT
 Hindsight: a trivia + calibration game
 Wed, 14 Oct 2026 16:58:40 GMT
 Some interesting internal documents from Anthropic litigation
 Wed, 14 Oct 2026 15:59:49 GMT
 PPA 11: The Basic Model
 Wed, 14 Oct 2026 15:03:34 GMT