Open weights remain irreplaceable for AI safety, argues response to Anthropic's anti-open-weights push
1a3orn · x · 2026-09-30
- Writing in response to Anthropic's latest not-very-oblique case against open weights, the author lays out how he currently thinks about open-weight AI models.
- Core claim: open weights have been deeply and irreplaceably useful for AI safety research, and shouldn't be regulated away.
- Follow-ups add concrete asks for Anthropic: equalize internal and external access much harder, treat bans from Claude as seriously as something beyond an ordinary SaaS ban, and make public commitments to that effect.
- A representative open-source-side counterargument to frontier labs invoking safety against open models.
More from AGI Musings
- Fei-Fei Li: Those fearing the intelligence explosion haven't considered the stupidity explosion — NaveenGRao · 2026-10-01
- Banning open weight models won't stop attackers, only enterprises — and that's the point — james_mtc · 2026-10-01
- Northwestern launches Economics of AI seminar series, Kleinberg kicks off Oct 6 — soumitrashukla9 · 2026-10-01
- McAfee: EU Tech Regulation Has Left European Productivity 20% Behind the US — amcafee · 2026-10-01
- Pedro Domingos: There are no software engineers anymore, we're all agent wranglers — pmddomingos · 2026-10-01
- Robin Hanson says AI will stay 'normal tech' with no foom; critics push back — pwlot · 2026-10-01