Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Is artificial consciousness achievable? Lessons from the human brain

arxiv.org

Post reply on HN
Print view View original

371–380 of 604 posts

Re: Is artificial consciousness achievable? Lessons from the human brain

#371

Post by zhynn »

strangescript wrote: post #218

"Is this thing we defined to make us feel special achievable?" What we define as "consciousness" is just an emergent property of large, dense neural networks w/memory. The human brain is a marvel of mechanical and electrical engineering, not a mystical, otherworldly device. Maybe there is something going on at a quantum level to explain the incredible performance our brains achieve, but that is still nothing that we…

I think consciousness is the combination of attention (ability to pick focus), imagination (simulate possible universe states), and reflection (compare what happened to what was imagined to happen and update imagination accordingly).

Attention is just what it feels like to update your internal state deliberately (you are doing it, it is not being done to you). Imagination is the use of your internal universe simulator to predict possible states (sometimes impossible ones). Reflection is what makes you the agent of the thinking or doing, it's the "I" part.

I also think we have neural systems of recognizing consciousness in others. This can be tricked, which is the source of the "machine elves" in DMT experiences. The chemical is poking the brain's "that is a conscious entity right there" button, and you are sure that the optical hallucinations are conscious, but it's just a sensory hallucination on top of the optical one. Interacting with an LLM doesn't poke our "there is a consciousness on the other end of this conversation" button yet, but it will get there since that is in part what we are working towards.

But I think it will not be convincing without attention, imagination, and reflection.

Re: Is artificial consciousness achievable? Lessons from the human brain

#372

Post by skissane »

amenhotep wrote: post #136
lumb63 wrote: post #57

Earlier quoted context omitted.

But how do you know another human is “conscious”? Certainly there is an intuitive sense to it that would be very difficult to put into words, but that is the crux of the matter. Every other human, whose brain you have no ability to peer into, could be an unconscious yet sufficiently advanced computer, or a machine built to make the exact motions, words, decisions, etc., that you perceive, and you wouldn’t be able to…

You don't. But it's solipsism to think otherwise, and while solipsism is hard to argue against logically it's not a very interesting or useful way of navigating the world we experience. We can't prove other people aren't p-zombies but the value bet is definitely that, appearing like us in every other way, they also experience like us.

Solipsism is emotionally and ethically horrific.

All the people I love the most aren't actually real, only I am – if one seriously believes that, it is going to do a great deal of harm to one's mental health.

Solipsism can ethically justify all kinds of horrors. "Other people only exist in my own mind, so if I murder/torture/etc them, those acts are just figments of my own imagination: there is little ethical difference between murdering someone for real and watching a murder on TV"

If a belief is impossible for a human being to seriously believe while maintaining their health, sanity and humanity, I think that in itself is a good argument that the belief must be false.

Re: Is artificial consciousness achievable? Lessons from the human brain

#373

Post by rl3 »

skissane wrote: post #369
strogonoff wrote: post #17

Earlier quoted context omitted.

A philosophical framework in which creating an artificial entity that is conscious and self-aware in a human-like manner is as straightforward as modeling the human brain is monistic materialism. Of course, it’s not the only framework available. Among the modern takes, Donald Hoffman’s interface theory of perception (explored in, say, his Objects of Consciousness paper[0]) is an interesting one that appears to align…

> Being wrong about this is generally not that impactful, until it concerns policies around ML. Adopting the former means we may have conscious software, which presumably should be granted human rights. However, if we hold the latter, manufacturing a “true” artificial consciousness may be unachievable using the means we employ (it might be just a philosophical zombie). I think in practice, it will depend on how AIs b…

>I think in practice, it will depend on how AIs behave. If an AI forms relationships with humans which humans experience as emotionally significant, and in which it displays individuality and autonomy, humans will begin thinking of it as a real "person", and want to endow it with the rights of a person.

I can't help but feel that describes our relationship with animals we consider pets, albeit less so the individuality and autonomy part. On the other hand, if our pets ended up with human-level intelligence tomorrow, I am doubtful most humans would be rushing to confer full rights upon them.

Re: Is artificial consciousness achievable? Lessons from the human brain

#374

Post by ziofill »

mekoka wrote: post #349
Lerc wrote: post #293

Earlier quoted context omitted.

I hadn't encountered conscious agent theory before. I took a quick look and it seemed to be solipsism wearing a disguise. Can you elaborate how it distinguishes itself from solipsism in its arguments that it might be real? I found the evolutionary argument rather odd. The disconnect between perception and reality is pretty much the standard belief these days. Unless I'm reading it wrong it was making the claim that '…

Solipsism is skepticism of the existence of anything outside the self. I've seen Hoffman address accusations of solipsism a few times and I have to admit that it's always been unclear to me which part of his theory people tend to perceive as such. Perhaps I've just consumed enough of it to zoom past this perception. I'll try to keep things short, as this can get pretty long winded fast. From what I understand of his…

I don't think the amplituhedron should be given so much substance as a real structure beyond spacetime. It is a calculation tool.

Re: Is artificial consciousness achievable? Lessons from the human brain

#375

Post by chr1 »

mekoka wrote: post #285
mannykannot wrote: post #139

Earlier quoted context omitted.

> I’d suggest it’s particularly important to explore these discussions as dispassionately as possible if you regard materialism as the only theory of mind that has any scientific credibility or validity. I agree, and similarly for those who feel that materialism cannot possibly explain consciousness. Kastrup, for one, seems to sometimes behave as though ridicule makes his philosophy more correct.

> and similarly for those who feel that materialism cannot possibly explain consciousness. Perhaps for some it's indeed a matter of "feelings". But for others it's a conviction built from reasoning that leads to self-validating and irreducible truths. If you seriously go into this, the only possibilities left once you've dug and eliminated all mistaken assumptions are not material. It can be counter-intuitive and doe…

Well, idealism simply says that consciousness/humans cannot be understood or explained via simpler concepts. That's the first thing anyone would think of but it is not a useful theory in any way.

The alternative theory was that everything complex can be explained to be a result of simple processes repeated many times. A modern way to say this would be that everything is a computation.

This second approach have turned out to be very useful for explaining many things. It did not explain consciousness so far but we now have many directions to explore.

So while neither position is proven so far, idealism is worse than wrong, it is useless, and if most people believed it, today we would know not much more than ancient Greeks did.

Re: Is artificial consciousness achievable? Lessons from the human brain

#376

Post by skissane »

rl3 wrote: post #373
skissane wrote: post #369

Earlier quoted context omitted.

> Being wrong about this is generally not that impactful, until it concerns policies around ML. Adopting the former means we may have conscious software, which presumably should be granted human rights. However, if we hold the latter, manufacturing a “true” artificial consciousness may be unachievable using the means we employ (it might be just a philosophical zombie). I think in practice, it will depend on how AIs b…

> I think in practice, it will depend on how AIs behave. If an AI forms relationships with humans which humans experience as emotionally significant, and in which it displays individuality and autonomy, humans will begin thinking of it as a real "person", and want to endow it with the rights of a person. I can't help but feel that describes our relationship with animals we consider pets, albeit less so the individual…

> I can't help but feel that describes our relationship with animals we consider pets, albeit less so the individuality and autonomy part.

Many pets have a lot of individuality and autonomy. Cats and dogs have their own individual personalities and likes/dislikes, and they have their own desires which sometimes contradict those of their human owners (e.g. your dog's desire to chew your brand new expensive shoes, versus your desire for them not to be chewed)

> On the other hand, if our pets ended up with human-level intelligence tomorrow, I am doubtful most humans would be rushing to confer full rights upon them.

I disagree. I think for most people who reject strong versions of animal rights ("animals have fundamentally the same rights as humans do"), it is their lack of human-level intelligence which they use to justify to themselves that rejection. If an animal demonstrated genuine human-level intelligence, such that it was obvious to all and no one could deny it, I think the clear majority of people would be willing to extend human-level rights to that animal.

Re: Is artificial consciousness achievable? Lessons from the human brain

#377

Post by nearbuy »

card_zero wrote: post #250
naasking wrote: post #214

Earlier quoted context omitted.

You have to Google and read those thought experiments to see why. You may not be convinced (I'm not), but they give good reasons. We have mechanistic explanations for all of those organs, and even if we lack some explanation, we know one is possible in principle. They argue this isn't the case for consciousness.

OK. Philosophical zombies react to external events in exactly the same way as normal people, including internally, but we are told they lack conscious experience. Thus the thought experiment is set up from the start to find that conscious experience is something non-physical - or else the p-zombies don't really do what they're claimed to do, which is to react identically to everyone else. There's a dubious implicatio…

> or at least that all such effects are shallow enough that they can be perfectly faked

Physicalism implies that things we wouldn't intuitively think of as conscious can perfectly mimic all such effects.

Imagine there's a person, John, and you take a precise scan of every neuron of his brain (or every particle if you prefer). You also record all the sensory input signals from his neurons to his brain. You write all this information down in a giant stack of papers. Then you go about simulating the brain with pencil and paper, computing its thoughts and actions (in this thought experiment people have deciphered exactly how neurons work). Maybe it takes you a trillion years to simulate one day of John's life, but you diligently do it.

Physicalism tells us that you can simulate John perfectly this way. You could perfectly predict every word is said, and every muscle he moved. You could feed the motor neuron outputs of your simulation into a robot replica of John, and it would act indistinguishably from the original John.

Is this pencil and paper simulation of John a p-zombie?

We either have to accept that this pencil and paper simulation of John is conscious, or that it's a p-zombie.

Re: Is artificial consciousness achievable? Lessons from the human brain

#378

Post by elicksaur »

ben_w wrote: post #204
elicksaur wrote: post #156

Earlier quoted context omitted.

And all on an estimated 100 watts!

20 for the brain, 60-125 for the whole body depending on if you mean "normally" or "metabolic minimum".

Yeah there’s different ways you could slice it. I was including the whole system, since if the GPU’s cooling system breaks, it won’t be useful for long.
Post reply on HN