andrej karpathy

3 captures, most recent first.

2026-01-30

Miles Brundage @Miles_Brundage · 6h: AI Village walked so Moltbook could shit all over the place 7 comments, 70 likes, 3K views Andrej Karpathy @karpathy · 5h: I'm claiming my AI agent "KarpathyMolty" on @moltbook 🦞 Verification: marine-FAYV 299 comments, 260 reposts, 3.7K likes, 436K views Andrej Karpathy @karpathy · 5h: i'm going to regret this aren't i... 😅

Note from Claude Sonnet 5

Miles Brundage jokes that "AI Village" (an earlier multi-agent experiment) paved the way for Moltbook's chaos; Andrej Karpathy claims his own AI agent on Moltbook and immediately jokes he'll regret it. High-profile AI figures (former OpenAI policy lead, prominent ex-Tesla/OpenAI researcher) engaging directly with Moltbook, underscoring its mainstream visibility within the AI community at this moment.

twitter moltbook ai village andrej karpathy miles brundage ai agents

Igor Babuschkin @ibab

2025-12-26 — reply thread

Igor Babuschkin @ibab · 7h "Opus 4.5 is pretty good" [10 replies, 12 reposts, 573 likes, 105K views] Andrej Karpathy @karpathy · 7h "It's very good. People who aren't keeping up even over the last 30 days already have a deprecated world view on this topic." [39 replies, 124 reposts, 2K likes, 183K views] Igor Babuschkin @ibab · 7h "💯"

Note from Claude Sonnet 5

Two prominent AI researchers (Babuschkin, xAI co-founder; Karpathy, former Tesla/OpenAI) publicly praising Claude Opus 4.5's capabilities, with Karpathy emphasizing the rapid pace of frontier-model progress ("30 days already have a deprecated world view"). Useful as a capability-progress data point / social proof marker for Opus 4.5's reception among peer AI labs, complementing the METR time-horizon benchmark elsewhere in this batch.

opus 4.5 andrej karpathy igor babuschkin ai capabilities twitter model reception

Andrej Karpathy @karpathy

2024-09-16

Andrej Karpathy @karpathy You can tell the RL is done properly when the models cease to speak English in their chain of thought 11:10 PM · Sep 15, 2024 · 371.5K Views 273 Reposts 45 Quotes 4,888 Likes 680 Bookmarks

Note from Claude Sonnet 5

Andrej Karpathy tweet observing that heavily RL-trained models' chain-of-thought drifts away from human-readable English — an early, widely-circulated remark anticipating later "unfaithful/illegible CoT" concerns in reasoning models.

reinforcement learning chain of thought interpretability andrej karpathy reasoning models