Hinton says AI agents may learn from thousands of lives at once
CurieuxExplorer · x · 2026-07-22
Geoffrey Hinton argues that AI agents could learn from thousands of lives at once, while humans only get to learn one life at a time.
The quoted clip also includes a more philosophical warning: you can back up files, but consciousness may not be something that can simply be copied. The post is essentially a short highlight from an Royal Institute talk, framed around the gap between human learning and what agents might eventually absorb at scale.
More from AGI Musings
- A post says the great American open-weights model matters more than the great American novel — arieljalali · 2026-07-22
- Opinion: Chinese Open-Weight Labs Appear to Catch Up Because OpenAI and Anthropic Stopped Releasing Models — Wide_Egg_5814 · 2026-07-22
- A 1964 Feynman talk is framed as the problem every AI lab still faces — HeyAmit_ · 2026-07-22
- A repost says adoption matters more than fighting over market share — srimisra · 2026-07-22
- A repost warns AI is collapsing language into one dominant voice — GabGarrett · 2026-07-22
- AI growth claims mask socialized risks, the post argues — SandraWachter5 · 2026-07-22