Sam Altman calls local inference a waste of resources; Dario Amodei agrees
beffjezos · x · 2026-11-01
Sam Altman argued that while many assume open-source models will let everyone run inference locally and shrink data centers, from a global allocation perspective it's terrible to have the world's limited DRAM supply sitting in personal computers used only 5% of the time. In a supply-constrained environment, compute should maximize cloud utilization. Dario Amodei voiced agreement with the take.
More from Companies & People
- Tesla Robotaxi teases 'spooky riders' roaming Austin on Halloween night — EricETesla · 2026-11-01
- Deepgram devrel team hosts rooftop Builder Lab hands-on hackathon — schwentker · 2026-11-01
- SF's ultra-rich will pay you to subscribe to their Substack — matt_slotnick · 2026-11-01
- Pedro Domingos mocks Anthropic's security plan: build dangerous AI, then block white hats — pmddomingos · 2026-11-01
- NZ RL lab Tacit discloses 7-figure pre-seed raise with full founder backstory — joecole · 2026-11-01
- Apple's new CEO John Ternus pushes back on 'behind in AI' narrative: 'We don't want to ship AI for AI's sake' — Hesamation · 2026-11-01