Scott Garrabrant comments on Why The Focus on Expected Utility Maximisers?

[Scott Garrabrant]

Note that I am not saying here that rational agents can’t have a utility function. I am only saying that they don’t have to.