This is a special post for quick takes by Daniel Kokotajlo. Only they can create top-level comments. Comments here also appear on the Quick Takes page and All Posts page.
Comment Permalink
8758
Yeah I think the Anthropic Consensus is disastrously false and is going to lead to misaligned Claude takeover.
If I agreed with it, I'm not sure what I'd do if I were in charge of Anthropic, probably I'd still do a bunch of different things, especially costly signals of genuine trustworthiness, but I admit I'd be a lot closer to their (predicted) behavior than I would in reality.
9