[roadside_picnic]

Can you just point me to the concrete examples (the most compelling examples in the book would work) where we can see "thinking" that performs something that is currently considered to be beyond the limits of computation?

I never claimed no one speculates that's the case, I claimed there was no evidence. Just cite me a concrete example where the human mind is capable of computing something that violates the theory of computation.

> "There are more things in heaven and earth, Horatio, than are dreamt of in your philosophy"

Fully agree, but you are specifically discussing philosophical statements. And the fact that the only response you have is to continue to pile undefined terms and hand wave metaphysics doesn't do anything to further your point.

You believe that computing machines lack something magical that you can't describe that makes them different than humans. I can't object to your feelings about that, but there is literally nothing to discuss if you can't even define what those things are, hence this discussion is, as the original parent comment mention, is "extremely boring".

[pegasus]

The kind of hard evidence you're asking for doesn't exist for either side of the equation. There is no computational theory of the mind which we could test "in the field" to see if it indeed models all forms of human expression. All we have is limited systems which can compete with humans in certain circumscribed domains. So, the jury's very much still out on this question. But a lot of people (especially here on HN) just assume the zero hypothesis to be the computable nature of brain and indeed, the universe at large. Basically, Digital Physics [1] or something akin to it. Hence, only something that deviates from this more or less consciously adhered-to ontology is considered in need of proof.

What keeps things interesting is that there are arguments (on both sides) which everyone can weigh against each other so as to arrive at their own conclusions. But that requires genuine curiosity, not just an interest in confirming one's own dogmas. Seems like you might be more of this latter persuasion, but in case you are not, I listed a couple of references which you could explore at your leisure.

I also pointed out that one of the (if not the) greatest physicists alive wrote two books on a subject which you consider extremely boring. I would hope any reasonable, non-narcissistic person would conclude that they must have been missing out on something. It's not like Roger Penrose is so bored with his life and the many fascinating open questions he could apply his redutable mind to, that he had to pick this particular obviously settled one. I'm not saying you should come to the same conclusions as him, just plant a little doubt around how exactly "extremely boring" these questions might be :)

[1] https://en.wikipedia.org/wiki/Digital_physics

[roadside_picnic]

> There is no computational theory of the mind which we could test "in the field" to see if it indeed models all forms of human expression.

I suspect the core issue here isn't my "lack of curiosity" but your lack of understanding about the theory of computation.

The theory computation builds up various mathematical models and rules for how things are computed, not by computers, how things are computed period. The theory of computation holds as much for digital computers as it does for information processing of yeast in a vat.

Evidence that human minds (or anything really) do something other than what's computational would be as simple as "look we can solve the halting problem" or "this task can be solved in polynomial time by humans". Without evidence like that, then there is no grounds for attacking the fundamental theory.

> What keeps things interesting is that there are arguments (on both sides) which everyone can weigh against each other so as to arrive at their own conclusions.

Conclusions about what? You haven't even stated your core hypothesis. Is it "Human brains are different than computers"? Sure that's obvious, but are the different in an interesting way? If it's "computers can think!" then you just need to describe what thinking is.

> how exactly "extremely boring" these questions might be :)

Again, you're misunderstanding, because my point is that you haven't even asked the question clearly. There is nothing for me to have an opinion about, hence why it is boring. "Can machines think?" is the same as asking "Can machines smerve?" If you ask "what do you mean by 'smerve'?" and I say "see you're not creative/open-minded enough about smerving!" you would likely think that conversation was uninteresting, especially if I refused to define 'smerving' and just kept making arguments from authority and criticizing your imaginative capabilities.

[pegasus]

In your previous comment, you seemed to have no problem grasping what I mean by "can computers think?" - namely (and for the last time): "can computers emulate the full range of human thinking?", i.e. "is human thinking computational?". My point is that this is an open, and furthermore fascinating question, not at all boring. There are arguments on each side, and no conclusive evidence which can settle the question. Even in this last comment of yours you seem to understand this, because you again ask for hard evidence for non-computational aspects of human cognition, but then in the last paragraph you again regress to your complaint of "what are we even arguing about?". I'm guessing you realize you're repeating yourself so try to throw in everything you can think of to make yourself feel like you've won the argument or something. But it's dishonest and disrespectful.

And yes, you are right about the fact that we can imagine ways a physical system could provably be shown to be going beyond the limits of classical or even quantum computation. "Look we can solve the halting problem" comes close to the core of the problem, but think a bit what that would entail. (It's obvious to me you never thought deeply about these issues.) The halting problem by definition cannot have a formal answer: there cannot be some mathematical equation or procedure which given a turing machine decides, in bounded time, whether that machine ultimately stops or not. This is exactly what Alan Turing showed, so what you are naively asking for is impossible. But this in now way proves that physical processes are computational. It is easy to imagine deterministic systems which are non-computable.

So, the only way one could conceivably "solve the halting problem", is to solve it for certain machines and classes of machines, one at a time. But since a human life is finite, this could never happen in practice. But if you look at the whole of humanity together and more specifically their mathematical output over centuries as one cognitive activity, it would seem that yes, we can indeed solve the halting problem. I.e. so far we haven't encountered any hurdles so intimidating that we just couldn't clear them or at least begin to clear them. This is, in fact one of Penrose's arguments in his books. It's clearly and necessarily (because of Turing's theorem) not an airtight argument and there are many counter-arguments and counter-counter-arguments and so on, you'd have to get in the weeds to actually have a somewhat informed opinion on this matter. To me it definitely moves the needle towards the idea that there must be a noncomputational aspect to human cognition, but that's in addition to other clues, like pondering certain creative experiences or the phenomenon of intuition - a form of apparently direct seeing into the nature of things which Penrose also discusses, as does the other book I mentioned in another comment on this page. One of the most mind bending examples being Ramanujan's insights which seemed to arrive to him, often in dreams, fully-formed and without proof or justification even from some future mathematical century.

In conclusion, may I remark that I hope I'm talking to a teeneger, somewhat overexcited, petulant and overconfident, but bright and with the capacity to change and growth nonetheless. I only answered in the hopes that this is the case, since the alternative is too depressing to contemplate. Look up these clues I left you. ChatGPT makes it so easy these days, as long as you're open to have your dogmas questioned. But I personally am signing off from this conversation now, so know that whatever you might rashly mash together on your keyboard in anger will be akin to that proverbial tree falling in a forest empty of listening subjects. Wishing you all the best otherwise.

PS: machines can totally smerve! :)