This deserves a detailed write-up, but the following list should be a start:
- The issue of outputs: if all there is is inputs and models, what are actions? In other words, how does one improve the chosen model’s accuracy?
- The place of the mind: TheOtherDave: all models exist in the mind, yet the mind itself is a model, does this mean that the mind contains models, yet the models also contain the mind? Which seems like a contradiction.
- The source of motivation and preferences: if everything is a model, what would be the reason to prefer one model over another and try to actualize it, and not just perform passion-free measurements?