I’m happy to see this sort of investigation. I wish there were 100 more investigations like this, so we could start to build up more of a thorough picture of the ways in which current AIs misbehave and reason about it. Thanks!
I’m happy to see this sort of investigation. I wish there were 100 more investigations like this, so we could start to build up more of a thorough picture of the ways in which current AIs misbehave and reason about it. Thanks!