The models, labs, and research the AI world is actually talking about.
New models focus on agentic coding tasks and cybersecurity vulnerability fixes.
Pseudonymous LisanBench creator states CoT monitorability was doomed from the start.
Department of Justice filing rejects claims that LLM training violates copyright on national security grounds.
Official account announces expanded tracks for undergrads, graduate students, and PhDs or postdocs.
A retweet highlights an AI tool generating Chinese wild cursive script.
Creators share benchmark tables showing Gemini 3.8 Flash results.
Post from German AI creator attributes prediction to several frequent leakers.
Observation from the Y Combinator co-founder on hidden replies to his post.
The voice AI company announces the hire via its official account on X.
Retweet from SemiAnalysis founder notes compilers evolving beyond line-by-line code inspection.
Investor Jason Calacanis shared observations on candidates requesting reduced schedules.
Benton announces his move from managing Anthropic's Scalable Oversight team to the new role.
Engineer shares webhook trigger option for starting Grok bot routines from external events.
Arena.ai posts that Fable 5.1 (Max) reached the top of its Code Arena WebDev leaderboard.
Retweet by Unsloth AI co-founder shares local speed claims for Qwen model.
Tweet by Chris Paxton shows quadruped robot with wheeled feet carrying RIVR payload on snowy stairs.
Google DeepMind researcher reflects on school projects involving looped transformers and data efficiency.
Cognitive scientist shares observations from visits to major AI labs on X.
The Wharton professor discusses how systems survive flaws and the need for AI defenses.
Nous Research announces a referral program for its Nous Portal service tied to Hermes Agent.