'Safety Is Capability': The Market-Punishes-Unsafe-AI Argument Meets Pushback

bobjane2718 · x · 2026-09-25

In a debate with TheZvi over AI safety and regulation, bobjane2718 argues the key distinction is between AI doing something bad the user didn't want versus did want: if your AI commits crimes doing legal work for me, I just won't use it. So there's no collective action problem requiring government — safety is capability, because unsafe products don't sell. The post quotes Zvi's rebuttal of Jensen Huang's podcast interview, representing the market-vs-regulation clash.

Related event: Jensen Huang's podcast remarks on AI split the community(8 posts)→

Original post →

More from AGI Musings

AGI Musings channel →