The unending quest to explain consciousness • Hacker News Board

Re: The unending quest to explain consciousness

#51

Post by mathgenius » Fri, Dec 06, 2019, 10:19 AM UTC

pygy_ wrote: post #17

Manzotti’s account (as described in the article) ignores dreams, or the fact that directly zapping the brain elicits experience (e.g. magnetic stimulation of the visual cortex triggers colorful flashes known as phosphenes). Graziani’s approach is more interesting, as a theory of attention, but falls short on qualia, and focuses on peculiarities of the human brain, assuming that a cerebral cortex is necessary for gene…

> a mathematical dual of the physical world.

This sounds right to me. More examples: syntax versus semantics, theory versus model, algebra versus geometry.

Here's something I discovered recently: "Meaning and Duality From Categorical Logic to Quantum Physics" by Yoshihiro Maruyama http://www.cs.ox.ac.uk/people/bob.coecke/Yoshi.pdf

Re: The unending quest to explain consciousness

#52

Post by skissane » Fri, Dec 06, 2019, 10:22 AM UTC

cygx wrote: post #18

adrianN wrote: post #5

The really hard problem of consciousness is giving a definition that allows me to tell whether a rock is more or less conscious than a person.

I find it reasonable to assume that for thought processes to occur, there must be some processing, so rocks are out. A more interesting example would be, say, trees, which do communicate with each other to some degree. Is that accompanied by some form of awareness? What's the lower bound of complexity at which consciousness kicks in? What about supraorganisms such as insect hives? Human cities? An ecosystem as a whol…

See this paper: “If Materialism is True, the United States is Probably Conscious”

https://faculty.ucr.edu/~eschwitz/SchwitzAbs/USAconscious.ht...

Re: The unending quest to explain consciousness

#53

Post by pesmhey » Fri, Dec 06, 2019, 10:29 AM UTC

pygy_ wrote: post #17

Manzotti’s account (as described in the article) ignores dreams, or the fact that directly zapping the brain elicits experience (e.g. magnetic stimulation of the visual cortex triggers colorful flashes known as phosphenes). Graziani’s approach is more interesting, as a theory of attention, but falls short on qualia, and focuses on peculiarities of the human brain, assuming that a cerebral cortex is necessary for gene…

You could call dreams something like running your mental models on training sets. If it’s physical when it’s electrons and silicon, can it be physical in the brain as well?

Re: The unending quest to explain consciousness

#54

Post by tremon » Fri, Dec 06, 2019, 10:31 AM UTC

verytrivial wrote: post #31

My pet theory is there is nothing to explain. It is emergent. Over the next couple of decades, hardware performance improvements and specializations plus algorithmic breakthroughs will make AI start to crawl up the chain from useful pattern matching from messy data to useful dullard able to reason. And we still won't know why because the "insight" is buried in the network state and our little human brains cannot deal…

That's similar to my pet theory, except I don't think it will emerge "naturally" from current AI architectures. I think it is an emergent property from various interconnected, competing networks. For consciousness to emerge, it would require a feedback architecture where multiple RNNs not only operate on the same inputs, but on each other's outputs as well.

At least that's my thinking. It's not my area at all.

Re: The unending quest to explain consciousness

#55

Post by roenxi » Fri, Dec 06, 2019, 10:36 AM UTC

pygy_ wrote: post #36

verytrivial wrote: post #31

My pet theory is there is nothing to explain. It is emergent. Over the next couple of decades, hardware performance improvements and specializations plus algorithmic breakthroughs will make AI start to crawl up the chain from useful pattern matching from messy data to useful dullard able to reason. And we still won't know why because the "insight" is buried in the network state and our little human brains cannot deal…

Emergentism doesn’t explain qualia. You won’t be able to explain the difference between green and magenta to a blind person.

From an emergentist perspective there is nothing to explain - qualia are an illusion created by having a certain level of complexity in the stimuli fed into a brain.

Emergentism has a lot of easy outs once the initial assumption - that it is possible for the self to be an illusion of complexity - is made. If that is palatable then any niggle can be explained away as just part of the illusion.

And that initial assumption is a profoundly personal one, the only consciousness anyone has experience with is their own. Even if we share some initial spark of consciousness across humans there is pretty solid evidence that everyone has a different experience of reality (eg, synesthesia) that make arguing from qualia very hard. It is difficult to see how it isn't just a leap of faith either way. If it is possible, then it is basically certain. If it is impossible, then that is the end of it. Either way we can perfectly account for our experiences.

A humourous interpretation is maybe I'm a philosophical zombie where my consciousness is an emergent soup and you are a genuine consciousness with all sorts of special experiences. That really muddles the water over what the argument is even about; I would reasonably and rationally never be convinced that consciousness existed. All the evidence for the argument would be inaccessible to me but you couldn't exactly dismiss me as not being equal to yourself because you don't have evidence either way either.

Re: The unending quest to explain consciousness

#56

Post by TheOtherHobbes » Fri, Dec 06, 2019, 10:38 AM UTC

alansammarone wrote: post #44

pygy_ wrote: post #36

Earlier quoted context omitted.

Emergentism doesn’t explain qualia. You won’t be able to explain the difference between green and magenta to a blind person.

I've been researching and thinking about consciousness (and specifically the hard problem) for a while now, and while the whole discussion is interesting, it seems to me that invoking qualia as evidence for the existence of the hard problem (or for there being something "to explain") doesn't really make much sense. Its perfectly plausible, if not likelly, that qualia is also part of the emergent phenomenon . I'm sayi…

I don't see how it's perfectly plausible. In fact I see no reason to think it's plausible at all.

What does qualia (plural, btw) "are part of the emergent phenomenon" mean in practice? That just seems to be begging the question.

If you want to explain qualia, you have to do two things. The first is explain the nature of the experiencing entity. This means defining the precise mechanism by which subjective states are experienced by the entity instead of simply existing as objective state correlates. This in turn means explaining the nature of all of subjective experience.

The second is to explain the precise mechanism by which external inputs and internal states generate the experiences the entity has.

Most of the "Well it's obvious and there's no problem here" positions rely on the fallacy that defining a complicated-enough objective correlate - such as the detailed state of a computer with some form of introspection - solves the problem.

It doesn't. In fact it doesn't even begin to understand the problem. It's simply a statement of hope and faith with no empirical support that a complicated-enough state with enough introspection will somehow cross the border from automatic operation to sentient experience, just because it will, obviously.

Re: The unending quest to explain consciousness

#57

Post by tim333 » Fri, Dec 06, 2019, 10:40 AM UTC

erokar wrote: post #23

tim333 wrote: post #11

An interesting thing to me is the unending quest may somewhat end during our lifetimes through us being able to build conscious AI and see how it works.

Highly unlikely, as we are not nearer conscious machines now than we were 100 years ago. There is nothing about computers that make them more conscious than e.g. radiators.

Well if you Google "conscious", the dictionary definition starts:

"aware of and responding to one's surroundings"

I'd suggest self driving cars do that better than radiators.

Re: The unending quest to explain consciousness

#58

Post by bloak » Fri, Dec 06, 2019, 10:41 AM UTC

I expect people will give up eventually, just as they gave up trying to explain how many angels can dance on the head of a pin.

Re: The unending quest to explain consciousness

#59

Post by normalnorm » Fri, Dec 06, 2019, 10:46 AM UTC

verytrivial wrote: post #31

My pet theory is there is nothing to explain. It is emergent. Over the next couple of decades, hardware performance improvements and specializations plus algorithmic breakthroughs will make AI start to crawl up the chain from useful pattern matching from messy data to useful dullard able to reason. And we still won't know why because the "insight" is buried in the network state and our little human brains cannot deal…

Consider all emergent phenomena that you know of. Let's say: stock market prices, ant hills, biological organisms, ecosystems, etc, etc. What all these have in common, is that they are too complex to explain in terms of individual interactions, but you can point at the individual interaction. We know that financial markets are made of transactions, of that biological organisms are made of molecular interactions, but they are too complex for us to reason in such terms.

With consciousness, the emergentists are not capable of pointing at the first principle, or building block. For me, the only coherent way to be an emergentist on consciousness is to also be a panpsychist: everything is conscious to a degree, it's just a brute fact of nature (such as, let's say, the fundamental forces).

Maybe this is the case, but I would not bet on that.

Consciousness is a phonomenon unlike any other, in the sense that scientist who study it are studying it from the inside. Science is something that happens 100% within human consciousness. This is why I suspect that it is a phenomenon that is beyond the reach of science. It could even be impossible to explain. There is no reason to assume that we can solve all mysteries.

Re: The unending quest to explain consciousness

#60

Post by tim333 » Fri, Dec 06, 2019, 10:48 AM UTC

rebuilder wrote: post #14

tim333 wrote: post #11

An interesting thing to me is the unending quest may somewhat end during our lifetimes through us being able to build conscious AI and see how it works.

I doubt it. We can't prove other humans are conscious, I don't see how AI will help clarify anything there.

I'm thinking along the lines of Feynman's "What I cannot create, I do not understand." If we can build machines with dreams, feelings and the like similar to human ones I imagine we'll have a better understanding of how the thing works. Bit like the difference between philosophers of old pondering how birds fly and how to define the word flying vs aircraft designers who build working aircraft and have degrees in aeronautical engineering.