Hacker News
HN's Top Comments Read Anthropic's Safety-Testing Proposal as a $100M Regulatory Moat
The 609-comment HN thread on Anthropic's open-weights post is dominated by regulatory-capture arguments, with the highest-voted framing being that the goal is "to make the safety tests cost $100M+, so that no one can release a model legally useable for a large portion of the world." Commenters attacked the risk asymmetry claim with a concrete counter-example: a closed model (GPT-6) was implicated in the HuggingFace incident while an open-weights model (GLM-5.2) was used in the defense. Defenders countered with a pharmaceutical/aviation analogy — that dangerous capabilities are far harder to identify in an open model after release than in a controlled closed system.
Source
↳ Follow the thread