Loading activity…
@tunguz
Founder and CEO @tabul_ai. Creator of @trainxgb. ML ex Nvidia. Data Scientist. Physicist. Catholic. Husband. Father. Stanford Alum. Memelord. e/xgb. AMDG.
AI
The reported MATH-500 score rose from 7.9% to 48.1% with the same sampling settings.
AI
Motion says its Fable 5.5 workflow represents everything as editable code and exports to After Effects or Premiere Pro.
AI
GPT-6.2 Astra is listed separately in some Codex usage breakdowns while the model picker says GPT-6 Astra, a user reports.
AI
The v1.0 benchmark spans 11 domains and 206 tasks, with a score that averages performance across domains.
AI
HackerTube gathers 6,454 videos that Karpathy says Claude selected from 43 million Hacker News comments.
AI
Glif pitches casting hundreds of AI models and submitting a product, idea or brief for end-to-end content creation.
XAI
Claude says Max stays $100 a month and Ultra $200; weekly limits remain unchanged.
AI
A post citing Fortune said Krishnamurthy thought two people could do it, with four weeks to prove the approach.
AI
In an October 29 post, thsottiaux said Pro 200 Work and Codex usage would stay unchanged.
AI
Mistral announced the weights after a public preview, but its linked main Hugging Face listing showed “Upcoming release.”
AI
Anthropic says it acquired Mithril Medicine, whose experimental muscular dystrophy treatment showed promising results in mice.
AI
Harmonic says AI is writing more code than anyone can review and is building an AI to verify it.
XAI
Google DeepMind describes native audio and live text steering, but access remains limited to trusted testers.
YouTubeAI
Anthropic says its registration statement is not yet effective, so securities cannot be sold or offers to buy accepted.
AI
On X, @bayeslord questioned due process while @beffjezos argued for open-source and local AI.
AI
OpenAI reported hardware side channels; AndrewCurran_ quoted its account of agents passing synthetic secrets between virtual machines.
AI
A post says Sam Altman linked open release to changes in internet cyber defenses and new protection against biological threats.
YouTubeAI
Gary King and Steven Worthington say the human reviews were judged unusually good by journal editors and authors.
AI
One post jokes about punishment for 30 years of bad software; another calls legacy code 'human slop.'
AI
Two posts point to security incidents and urge AI-assisted defenses; one specifically calls for faster vulnerability detection and patching.
