OpenAI researchers imagine the counterfactual world where o1/RLVR was never discovered

willdepue · x · 2026-09-22

antimatter15, OpenAI's willdepue, and Nat Friedman riff on a counterfactual: had o1/RLVR never been discovered, voice interactions would have matured earlier, Jevons paradox would have arrived sooner, chat would be funnier and more "wordcel," and AI might already be an ad-driven consumer business. willdepue adds that many ideas were left on the table in GPT-4 as the field moved too fast. An interesting insider counterfactual discussion.

Related event: OpenAI Staff Question Chat Format, Imagine a World Without RLVR(4 posts)→

Original post →

More from AGI Musings

AGI Musings channel →