Posted: Nov 14, 2017 7:12 pm
GrahamH wrote:Tse rejects computational theory of mind on the basis that brains are not algorithmic, not UTMs, but goes on to talk about criteria systems handling large numbers of inputs in parallel that handle non explicit rules. Now that the sort of thing deep learning systems are getting very good at so his rejection may be premature and overly simplistic.
Cool. When I listen further to him, I'll watch out for that. I'll also be wondering if he's essentially proposing some sort of two (or multi) stage free will process in which the randomness (via 'noise') you say he cites provides a decoupling from determinism, follwed by a stage in which the agent makes choices.
If so, I'll be wondering how that isn't just shunting the key problem, unresolved, into the second (or final) stage, which is what I think is the shortcoming of the QM two-stage models.