AI Daily · Nov 2, 2026 · Monday
BAD INFLUENCE
Karpathy's HackerTube curates videos from 20 years of Hacker News comments
Rohan Paul relays Andrej Karpathy's account of HackerTube, a site he built in about two hours with Claude Code and Opus 5.5 on Medium effort. Claude filtered 43 million Hacker News comments over 20 years down to 6,454 recommended videos. The author says he only supplied a Google Cloud API key, a VPS and a domain name, and that the site is available at hackertube.org.
Model releases & updates
SemiAnalysisBlog1 more sourceSemiAnalysis launches GDPx, a benchmark covering high-value frontier AI domains
SemiAnalysis announces GDPx, a benchmark of agentic tasks across domains it says will generate the most frontier AI revenue, including chip design, cybersecurity, biology, and finance. Its tasks run in realistic environments built with 14 specialized data partners, and it says the top model scores 12%. The article contrasts this design with OpenAI's GDPval, which it says gives models all context in the prompt and relies on rubric grading.
Industry news
Anthropic restricts Claude Code from cloning games as AI reshapes game piracy
Anthropic added instructions to Opus 5.5's system prompt that discourage Claude from recreating specific existing games, after an X post drew attention to the change. The article says the restriction covers only Claude, so users can route game-cloning tasks through other models, and it has also blocked legitimate original game development. It frames AI-driven cloning and console jailbreaks as a possible 'Napster moment' for the games industry, citing an indie studio that cancelled a public demo over fears of AI copying.
The DecoderNews2 more sourcesOpenAI merges Chat and Work in ChatGPT, sharing one token budget from December
OpenAI is merging the Chat and Work modes in ChatGPT, according to an updated support page. Starting in December, Plus, Pro, and Business plans will share a single token budget across ChatGPT and Codex. Chats with Instant models will not count toward that budget, but heavy use on cheaper plans will likely mean lower-quality results.
Tips & takes
Mustafa Suleyman@mustafasuleymanXMustafa Suleyman argues against building conscious superintelligence
Mustafa Suleyman says he has seen no evidence that AI systems are conscious or deserve welfare, and that developers should not build conscious AI. He says that if consciousness emerged in training, development should stop until those properties could be designed out. He also says he disagrees with Anthropic's approach while praising its transparency.