Gary Marcus

This. I'd like to add: bias which can often be learned, which humans did. I don't think that humans were born with hard-coded reasoning capabilities. We learned them from data (observations and interactions). When we humans deal with the problem above, we use years of experience.

you might want to read Bonatti’s recent Science paper or Gervain’s newborn work that follows up my infant rule work. or to think about what evolution does to brains over time. (see my book The Birth of The Mind.)

Ok, but how does that change the argument? Evolution is machine learning with a longer time span, e.g. an optimization process. It is still machine learning. The bias is learned by evolution.

i addressed this in my arxiv on alphago and innateness

If I am not mistaken, your paper responds to my first reply (we are not born blank), but not to my second reply, which says that "innateness" is not a prior appearing from nowhere. What you call "innate" is the result of an additional learning process, performed by evolution.

actually, i address exactly that

IIRC your paper points out that such args dilute the meaning of learning. I don't agree with this* but I do agree arguments that evolution learned it are non-sequiturs and completely irrelevant. Humans don't start from blank states; it doesn't matter what installed those biases

The question was whether ML can address the problem above. If evolution was abled to learn these biases, why wouldn't ML be able to learn them? In any case, humans have them, and if you rule out that a higher being put them there, some form of learning created them.