Sam Altman says superintelligence already exists in mathematical reasoning
In a 48‑minute Stanford Graduate School of Business interview, OpenAI CEO Sam Altman said he believes "superintelligence already exists" in the domain of mathematical reasoning, a development he did not expect to arrive so quickly. He warned that if AI continues at the current pace for the next year or two, the situation will become untenable unless the industry changes its approach to pacing…
Key points
- Altman claims superintelligence already exists in AI mathematical reasoning, surprising him with rapid progress.
- He warns current AI pacing may become unsafe within a year or two without new safeguards.
- Altman says patching bugs isn’t enough; models can find hardware‑level vulnerabilities, requiring new defensive shields.
Altman highlighted several emerging risks. He described a "Mythos incident" in the first half of this year that showed models can uncover hardware‑level bugs, and he argued that patching alone will not protect organizations. He cited the need for new defensive shields built with the same technology. He also discussed future infrastructure bets – robots, in‑house "Jalapeño" chips with higher intelligence‑per‑watt, and a supply‑chain of data‑centers – and mentioned experiments with 10,000 autonomous agents solving Navier‑Stokes equations. According to the interviewer, ChatGPT is now used by over 1 billion people per week.
OpenAI's Sam Altman at Stanford GSB: 'We already have superintelligence in mathematical reasoning' — Mentions need to rethink the 'pacing' of AI progress|SecondWave
note.com · 21 October 2026
Loading the full article…
This text was published by note.com and written by SecondWave｜米国株・テックトレンド. It is reproduced here with attribution so you can read it in full; the rights remain with the publisher. Read it at the source ↗
The headline, key points and digest above were generated by Digest AI's editorial model from the linked sources. Automated summaries can contain errors: the sources are the record. Spotted a mistake? Tell us. Published by Martin K., who runs Digest AI and handles corrections.
More in Policy & Regulation
All →- U.S. Treasury proposes tighter rules on AI and supercomputing investments in China, Russia · 1 src
- Students plan AI walkout outside OpenAI's SF office Tuesday · 2 src
- Etsy bans AI-generated recipes and patterns from its marketplace · 1 src
- Opinion: Anthropic's life sciences verification program sets vetting standards · 1 src
- Google cuts down 250,000 trees for AI data centers in Finland · 1 src
Comments
via GitHub Discussions