>> So in this sense a minimal set of a-priori concepts are required to be built-in, or else we couldn't learn anything at all.
I don't disagree with that at all. I'm pretty convinced that, as humans, we can learn and invent all those things we have because we have strong inductive biases that guide us towards certain hypotheses and away from others.
Where those inductive biases come from is a big open question, and I'd be curious to know the answer. We can wave our hands at evolution, but that doesn't explain, say, why have the specific inductive biases we have, and not others. Why do we speak human languages, for example? Why is our innate language ability the way it is? Intuitively, there must be some advantage in terms of efficiency that makes some inductive biases more likely than others to be acquired, but I get tired waving my hands like that.
I'm not convinced that all that absolutely requires a body, either. I think it's reasonable to assume it requires some kind of environment that can be interacted with, and some way to interact with the environment, but why can't a virtual environment, like a computer simulation of reality, provide that? And it doesn't have to be the real reality, either. A "blocks world" or a "grid world" will do, if it's got rules that can be learned by playing around in it.
I don't disagree with that at all. I'm pretty convinced that, as humans, we can learn and invent all those things we have because we have strong inductive biases that guide us towards certain hypotheses and away from others.
Where those inductive biases come from is a big open question, and I'd be curious to know the answer. We can wave our hands at evolution, but that doesn't explain, say, why have the specific inductive biases we have, and not others. Why do we speak human languages, for example? Why is our innate language ability the way it is? Intuitively, there must be some advantage in terms of efficiency that makes some inductive biases more likely than others to be acquired, but I get tired waving my hands like that.
I'm not convinced that all that absolutely requires a body, either. I think it's reasonable to assume it requires some kind of environment that can be interacted with, and some way to interact with the environment, but why can't a virtual environment, like a computer simulation of reality, provide that? And it doesn't have to be the real reality, either. A "blocks world" or a "grid world" will do, if it's got rules that can be learned by playing around in it.