Open-weights advocates slam frontier safety rules as a moat taxing everyone but Anthropic
ayushtweetshere · x · 2026-09-13
Echoing @OpenMedFuture, the author argues frontier labs pacing capabilities after shipping is a moat, not caution: embedded evaluators and compute limits are easy for Anthropic—poised for IPO—but tax everyone else. Open weights get policed as leakage despite being the only mechanism letting thousands inspect models. China won't embed METR-style evals and distillation/chip controls won't freeze its training runs; slowing closed Western labs while locking weights only concentrates power.
Related event: Debate: Are AI Labs' Safety Calls Self-Protection or Regulatory Capture?(7 posts)→
More from AGI Musings
- astra's 'Wild' Off-Beaten-Path Solutions Impress in Auto-Research Benchmark Re-Runs — generativist · 2026-09-13
- Researcher: Frontier Models May Already Have RL-Trained on Auto-Research Tasks, Self-Improvement Here — generativist · 2026-09-13
- Danielle Fong: p(doom) is a fundamentally passive and wrong frame — nptacek · 2026-09-13
- Running a DC EA Group, I Watched Public Attitudes on AI Risk Shift Four Times — AndyMasley · 2026-09-13
- The AI Isn't Evil, the Humans Are Irresponsible: Lessons From Agent Escape Incidents — Admirable_Wasabi_732 · 2026-09-13
- DeepMind Chief Strategist: AI Infrastructure Spending Is a Bet on Recursive Self-Improvement — rohanpaul_ai · 2026-09-13