Open Weights Still OK, but Pressure the Inference Layer? AI Circle Debates
willcb · x · 2026-09-13
Joshua Saxe says he long believed openness and diffusion was the right strategy for a good cyber equilibrium in the AI era, but is now reconsidering. His loosely held view: keep supporting open weights for the foreseeable future while preparing heavy pressure on inference providers (Together, Fireworks, and closed labs). Discussion also notes the reported Stripe acquisition of OpenRouter looking smarter by the day—"an unlikely savior."
Related event: Cybersecurity Experts Debate AI Offense-Defense and Inference-Layer Rules(4 posts)→
More from AGI Musings
- New article: Intelligence Has a Speed Limit — why RSI can't run as fast as you like — NathanpmYoung · 2026-09-13
- 25-year cybersecurity veteran calls AI safety narrative and METR authority 'dangerous and deceptive' — basedjensen · 2026-09-13
- Ex-OpenAI exec Zack Kass: software got cheap because nobody could gate it; housing and healthcare got a permission layer — victor_explore · 2026-09-13
- Dan Jeffries Rips AI Doomers: Extinction Scenarios Break Physics and Common Sense — Dan_Jeffries1 · 2026-09-13
- Prediction: NS problem-solving replicable on consumer hardware at home within 10 years — int_16h · 2026-09-13
- Robotics academics in despair as Astra, Fable and Muse zero-shot benchmarks, says NYU professor — LerrelPinto · 2026-09-13