Posted: Aug 26, 2017 7:31 pm
romansh wrote:scott1328 wrote:
I don't believe the alphaGo has the capacities you attribute to it. It has no model. It does not evaluate outcomes. It is,at best, a situatuation reaction machine; an expensive thermostat.
Better, that you should have asked me about my dog's free will.
I don't know whether it has a model, I suspect it does and as it plays games it updates it model.
Does it compare different outcomes and choose amongst them? It by definition must do.
I agree it is an expensive thermostat ... but then so what?
What it almost certainly doesn't do is know it is doing any of that. But maybe that would only a more elaborate game than Go.