AI labs' job posts reveal the next breakthrough: endless data harvesting and RL environments
DuaneJRich · x · 2026-11-22
By cross-referencing job posts from OpenAI, Anthropic, and Mercor with benchmarks models still struggle on, one analysis concludes that smarter models so far mostly come from endless data harvesting.
- Anthropic has a computational chemistry role explicitly about building RL environments to teach Claude reaction prediction and synthesis planning, plus an antibody scientist to test AI-designed candidates and feed results back to researchers.
- OpenAI's Science data role says its current focus is biology and chemistry; both labs are hiring people to build environments.
The poster adds that data harvesting has already produced remarkable intelligence, but an operational and financial "law of gravity" acts as an inescapable friction — a natural answer to "please slow down."
More from AGI Musings
- AI poster teortaxesTex 'makes it to the boomersphere' via Ross Douthat — teortaxesTex · 2026-11-22
- Self-replicating ambient AI actuators as techno-capitalism's archetypal end — LesaunH · 2026-11-22
- Compilation of self-proclaimed AI experts failing to foresee AI's rapid progress — Chris_Armstrong · 2026-11-22
- ngxson essay: Code was my second mother tongue — now everyone speaks it — ngxson · 2026-11-22
- AI skeptic: market crash could permanently erase AI stock premium, then a supply glut hits — teortaxesTex · 2026-11-22
- "Go to a Family Reunion and Ask About AI": A Wry Take on the Bubble Debate — PurzBeats · 2026-11-22