bargo comments on Curiosity as a Solution to AGI Alignment

[bargo]

Solomonoff Induction and Machine Learning. How would you formulate this in terms of a machine that can only predict future observations?