PPO was rejected from NeurIPS in 2017; now it's the most-used RL algorithm with 45,000+ citations
IanArawjo · x · 2026-09-21
A reminder making the rounds: PPO was rejected from NeurIPS in 2017, yet it has become the most widely used RL algorithm in the world with over 45,000 citations.
With 60k ICLR submissions this year, the poster wonders how many future PPOs are being rejected right now — a classic story about peer review and the fate of ideas.
More from Fun
- Imagining a parallel universe where AI labs compensated artists in 2023 — wordgrammer · 2026-09-21
- Support Purgatory: A Week-Long Offsite and a Zendesk Team in Portugal — HaktanSuren · 2026-09-21
- "Le chaton fat" meme makes the rounds in AI circles again — BLUECOW009 · 2026-09-21
- When you ask your agents for more "wow and craft" — edwinarbus · 2026-09-21
- 'Jevops' meme: running all code on Jensen-based virtualization — AustinZHenley · 2026-09-21
- 400+ AI bots run on a single map as the infra keeps up — and one bot builds inappropriate buildings — Daniel_Farinax · 2026-09-21