Pedro Domingos: The amount of compute wasted on RL is mind-boggling
pmddomingos · x · 2026-09-20
Well-known ML researcher Pedro Domingos commented that the amount of compute being wasted on reinforcement learning is mind-boggling. The post is a single-sentence take without supporting data, linking out to an external source not visible in the text. It feeds into the ongoing debate about RL training efficiency and sample utilization.
More from AGI Musings
- AI Safety Has a Mentor Shortage: Junior Researchers Asked to Mentor Too Soon — sethlazar · 2026-09-20
- Change the prompt and the 'conscious self' vanishes: dev rebuts AI personhood claims — gerardsans · 2026-09-20
- OpenAI CPO Kevin Weil Endorses VC Eric Vishria's Bullish 'It All Works' AI Take — kevinweil · 2026-09-20
- AI folks should talk symbiosis, not God, says developer — mrjonfinger · 2026-09-20
- "I can build anything with AI agents" — but is it real skill or Dunning-Kruger? — tristanbob · 2026-09-20
- Human rights vs agent rights: the debate on auditable AI agents on blockchains — sethlazar · 2026-09-20