... hurt: a mass email to 50k users, or a migration you can only roll back after downtime. "Reversible on paper, ruinous in practice." His version is that the gate should key on reach, not on reversibility. Not "is this safe" but "how far does this reach if I'm wrong." And the part that stuck with me: the agent is the worst possible judge of its own blast radius, because it does not know there are 50k people on that ... ... when unsure rule into something testable. The important missing piece is a safe

... useful. Comment from Navanita Devi(@navanita_devi): Dedy's point is interesting, the opacity problem is real. Most people conflate "AI having access" with "AI having unchecked control." They're very different things. Read-only access vs write access, logged actions vs