OpenAI hits automated research intern goal, agents solve Navier-Stokes in 88 hours
btibor91 · x · 2026-09-13
OpenAI this week
- Says it reached its automated research intern goal and is progressing toward an automated AI researcher by March 2028.
- Paused RL training on deployment-bound models after the Hugging Face incident; an investigation found Tristan Buckmaster's Codex prompts could not have influenced the system and no user data was accessed.
- Chief Scientist Jakub Pachocki's essay An Alien Mind argues no lab has solved alignment and monitoring well enough to keep scaling responsibly at maximum speed for much longer.
- Published a solution to the Navier-Stokes Millennium Prize Problem, produced in 88 hours by 10,000 coordinating agents running a new internal model more capable than GPT-6 Astra.
- Called for mandatory capability-based national regulation.
Anthropic this week
- Cyber incidents and a development slowdown headline its news.
The week packs automated-researcher timelines, an alignment-pessimist essay, and a Millennium Problem solve into one snapshot of the acceleration vs. slowdown debate.
More from Companies & People
- Stack Overflow responds to 99.5% question drop: enterprise pivot and agent platform — pchandrasekar · 2026-10-30
- Every's inaugural Thesis: 2027 conference to explore human work after automation — every · 2026-10-30
- Applied AI interview lesson: stop leading with what you aren't, sell the problems you solve — brandon_galang · 2026-10-30
- AI company tier list puts NVIDIA, Microsoft and OpenAI in the top S+ tier — agihippo · 2026-10-30
- OpenAI to shut down Prism, its Overleaf rival, on January 1, 2027 — aniketapanjwani · 2026-10-30
- Former OpenAI artist-in-residence's bronze work lands the cover of Pearson's intro art textbook — artBoffin · 2026-10-30