Matthew Barnett comments on Predictable updating about AI risk

[Matthew Barnett]

Definitely. I don’t think it makes much sense to give people credit for being wrong for legible reasons.