Isn't this sort of the other way around? It's called Machine Learning because it's at least somewhat analogous to, y'know, learning. It just seems weird to me to apply psychological concepts onto systems that were deliberately designed in order to replicate psychological effects as if that's something new. Of course maximizing optimization functions is a conceptual way to think about decision making; that's why started using optimization in decision-making algorithms in the first place.

Yeah I guess you could make the case for it being the other way around. Although, I think that a lot of ML concepts have groundings in optimization, probability, and a bunch of other mathematical concepts as opposed to concepts that wanted to replicate those psychological effects.

I think that there is much hype on ai and neuroscience. Actually we know many thinks of the brain as organ but we know really little about mind think and consciousnes. There isn't scientific evidence that consciousness have a material origin. And actually all the air we do is based on weak ai

Interesting post. I had written something similar a couple years ago as well. Nice to see someone else see the same patterns.

http://wp.me/p3SyBR-4l

I thought the regularization analogy was kind of a stretch, but otherwise, everything else was on point!