Roman Leventov comments on Why The Focus on Expected Utility Maximisers?

[Roman Leventov]

Reward is not Necessary: How to Create a Compositional Self-Preserving Agent for Life-Long Learning