Hacker News new | past | comments | ask | show | jobs | submit login

I wouldn't say that's really a problem, because there's a perfectly fine scientific explanation of why it doesn't (couldn't) "capture" this. The experience of color or taste is triggered when information is fed to the brain through specific pathways. The pathways that are activated when we read about something are different and don't reach the same places, so any information you acquire through reading and cognition is necessarily going to be "tasteless". Because of that, we can't possibly "connect" our knowledge of the process that causes the qualia of taste to the qualia of taste, but that's just a limitation of the brain's wiring, not a fault in the model itself.



> the experience of color or taste is triggered when information is fed to the brain through specific pathways

The question is why the experience is triggered. Why does moving atoms at specific positions trigger the perception of color or pain? Current physics doesn't have an answer for that, even in theory, assuming we can see exactly what happens in the brain.


The thing is, though, that even if "physics" did have an answer to this, it's not a given that we would understand it. Our incapability to wrap our head around reductionist explanations of qualia does not entail that they are inaccurate, only that they fail to satisfy us. I mean, let me put it this way: what do you expect from an answer?

For example, in order to understand why atoms moving in certain ways produce a taste, do you need the explanation to trigger some experience or memory of a taste? If so, you will never understand.

Or do you expect the answer to be short? What if qualia tend to be as complex as the brains that experience them (for holistic reasons), so the proof for our own qualia spans trillions of pages? I think that's actually quite probable, but you'll never understand in that case either.

The space of naturalistic explanations of things is humongous and we're already mass producing objects so complex no single individual could hope to understand every aspect of them. If you're going to start with "moving atoms at specific positions" in order to understand a defining aspect of a system made out of hundreds of billions of complex switches connected in obscure ways, let me tell you that you're going to need several Libraries of Congress before you're done.


What I expect from an answer to the hard problem of consciousness is a basic high level explanation for how the moving of atoms in certain ways can actually give rise to qualia. You seem to be starting with the presumption that it does. But I'm not convinced that the moving of atoms alone can possibly give rise to qualia, because our knowledge so far about how atoms move seems completely different from our experience of qualia.

On the other hand, I don't understand all the details of what goes into a Macbook Pro, from network protocols to graphics rendering, but I have a basic high level understanding of how the things it does are possible, and it doesn't take a Library of Congress to explain that. My high level knowledge of how atoms move provide sufficient assurance and understanding of how a Macbook Pro experience is created. That is simply not true for qualia.


> My high level knowledge of how atoms move provide sufficient assurance and understanding of how a Macbook Pro experience is created.

I don't think it does per se. In a world where computers didn't exist you would probably not be able to say whether a MacBook Pro is a thing that could possibly exist or not just from fundamental knowledge of physics. If I gave you the most low-level possible description of physics, it'd probably take you some time to figure out whether it supports solid macroscopic objects. If I just give you the rules to the Game of Life cellular automaton, it's far from obvious that it supports replicators, let alone that it's Turing complete.

Most of these things are only obvious in hindsight, and the understanding is made easier by the lack of cognitive dissonance: we know we made them using certain principles, so of course it works.

> That is simply not true for qualia.

I don't know about that. My own high level knowledge of physics does provide me with sufficient assurance and understanding of how qualia happens. Perhaps I'm wrong, but I do feel like I understand it: what defines or distinguishes, say, my qualia of red, is how it primes my brain into thinking about red things, how it changes my mood, how it behaves relative to every other concept, which makes it about as complex as my entire brain would be. I also imagine that there is a continuum of "qualia" such that basically any physical system has some, but they are so trivial we don't make the connection.

Again, that could be nonsense, but it makes perfect sense to me -- about as much as a MacBook does. The hard problem that you see flat out does not exist to me, and presumably the reason is that we are satisfied by different explanations.


> Perhaps I'm wrong, but I do feel like I understand it: what defines or distinguishes, say, my qualia of red, is how it primes my brain into thinking about red things, how it changes my mood, how it behaves relative to every other concept, which makes it about as complex as my entire brain would be.

Over here it seems like you're defining consciousness by its relation to other objects. In other words, you're reducing consciousness to its visible outputs, such as behaviour, mood, etc. If that's all consciousness is, I fully agree with you that we have sufficient knowledge of physics to explain it. But that's not what I'm referring to. When I talk about consciousness, I'm referring to something that is completely independent of its outputs. Is it not conceivable that a man who is missing both his arms and legs, mute, blind, and deaf, can still have a conscious ego? So what is that experience? This is the question which has no parallel in our knowledge of how atoms move.


> I'm referring to something that is completely independent of its outputs. Is it not conceivable that a man who is missing both his arms and legs, mute, blind, and deaf, can still have a conscious ego?

The brain has many internal inputs and outputs, even more of them than external ones. Even if you imagine a completely isolated brain that can't see, hear or move, it's still exchanging signals internally. For instance, such a brain could fabricate a dream world, create (possibly nonsensical) images to feed to its visual cortex directly, and so on. Its conscious experience would be whatever it builds for itself, operating through these internal input and output channels, shaped by random noise and whatever memories it may have of real things. It's what dreaming is, basically. That is consistent with my view.

If, on the other hand, you ask me to imagine a mute, blind, deaf, paralyzed man who also has his brain cut up in such a way that it can't talk to itself, then no, I do not think that it is conceivable that the man has a conscious ego.


It seems once again that you're defining the qualia in terms of some input/output formula. But that's missing my point; I used "external" inputs/outputs merely as an example. At what point do those inputs and outputs, whether they be internal or external, become qualia? In other words, from where does the "ability to experience" the inputs and outputs come from? Does the mere fact that the brain receives input make it able to experience the inputs?


I never brought up inputs and outputs, though, you did, and ultimately I feel it's kind of a strawman of my position.

A brain is made out of billions of neurons, each of which has its own inputs and outputs, and so does each part of each neuron, down to the atomic level. Every physical entity is a kind of IO machine, so in a sense, any physical definition I may give of qualia will reduce to IO patterns. That is, however, missing the bigger picture, which is the structure of the system: how the information flows within it, and how certain conditions modify that flow. That's where the meat is.

In my mind, consciousness and qualia are structural properties. When experiencing a qualia, the information flows in a particular way in your brain, like a mode of operation, and that's what the qualia is.

I think of it this way: you experience "anger" when your brain is put in an "anger mode", which taints everything it sees or does -- anger is not an "input" or an "output", it is a "mode" that your whole brain is in. It makes you "different", although not in a way that makes you lose your sense of self (well, usually). And in a lesser way, when you see red, your brain enters a "red mode", when you eat nougat it is put in a "nougat mode", and when you think about red or nougat your brain also enters a "recall" version of red or nougat mode that has similar properties.

So what happens is that your brain can enter billions of subtly different "modes" that correspond to different stimuli (real or not). Each of these "modes" makes you think a bit differently, in a way that's adapted to the stimuli you received. These are the "experiences" it can have: they are not inputs or outputs, they are ways of thinking, triggered by stimuli or memories or sometimes random noise.

I guess I would say a qualia is a bit like a protein fold, it's a functional reconfiguration.


I understand that you view qualia as structural properties, and can map trillions of physical brain-configurations to trillions of subjective brain-states. However there is still a massive discrepancy between our subjective experience of such states, and the physical configurations that are claimed to represent it - which is why it is not a satisfactory answer to the hard problem of consciousness for many people.

The 2 big discrepancies I see are this: 1. We have the subjective perception of free will: I can choose to write these sentences or not. Where does that intention originate? If our consciousness is just neuronal computation, then it is completely deterministic or probabilistic. That is entirely contradictory of most people's experience, because we can choose what we want to do, and do things that strongly contradict our "mood". Some people can even go further to consciously modify their brain states. If you claim that is also just a product of complex structural properties, you get into an infinitely regressing claim.

2. Despite being able to create incredibly complex connected structures, arguably even more complicated than a single person's brain, such as the Internet, we have no evidence of such an entity having a subjective feeling of consciousness. There is simply no parallel in our current knowledge for how subjective experience can arise from system configurations. You seem to be simply assuming that an experience is such a configuration, but we have no evidence of that. If such a configuration could be consciousness itself, then we have examples of other entities which are connected in similarly complex ways, yet we don't believe they are concious. This argument is explained here: http://www.faculty.ucr.edu/~eschwitz/SchwitzPapers/USAconsci...


1. I find that one fairly easy to explain (easier than qualia, even). When we weigh decisions, we can use an internal model of ourselves to support our thought process. We can imagine ourselves doing X, evaluate the consequences a bit, assess how it feels, then imagine ourselves doing Y, and so on. Considering that our intuitive understanding of "I can do X" corresponds to "I can imagine my self-model doing X, and I don't see any external factors that would stop me," then our ability to evaluate multiple courses of action directly entails an impression of free will.

Now, this "free will" only corresponds to epistemic non-determinism about ourselves: "for all we know", we can do X, and we can also do Y, because if we knew we couldn't, we wouldn't be evaluating these possibilities. So in a sense, it is our self-model that has free will: it is the model that we imagine in various circumstances, but the model isn't the real thing. So our mistake is to assume that this property of the model is also a property of the real thing, i.e. that our epistemic uncertainty translates into some kind of metaphysical non-determinism.

I would expect that any decision system that can construct hypotheticals about itself would have an impression of free will, or a qualia of free will. They may not necessarily be able to express it, or to feel it as vividly as we do, but they would have it.

2. I know the point you are trying to make, but where you see a reductio ad absurdum I just see a lack of imagination. Our usual understanding of consciousness is very, very, very deeply anthropomorphic. Neither the Internet nor the USA are systems that have a human or animal consciousness, that much is clear. "But what about consciousness in general?" Man, I'm sorry to be blunt here, but you wouldn't know it even if it hit you in the face. We're human. We have human consciousness. Our idea of consciousness outside of the exact thing that we have is: fuck. all.

We can start with Leibniz's thought experiment they mention in the article, where a gigantic brain is built out of gears and pipes. Or the similar thought experiment about the "China brain" which emulates a human brain by using one person to emulate each neuron. To me, it is clear that both of these things would, in fact, be conscious. I see no good reason to think otherwise and I am perfectly content biting that bullet. I have zero reservations, even intuitively.

Now, regarding something like an "USA consciousness". Personally, I do not think this is bizarre. It's only bizarre under an anthropomorphic view of consciousness, or to put it in an other way, it's only bizarre if you believe that conscious entities ought to be relatable (an absurd expectation). But if you use the criteria I've given, you can find several things that would correspond to "USA's qualia". For example, 9/11 caused large structural changes that you could interpret as the USA feeling a qualia of fear and/or anger. A new episode of Game of Thrones may also be a qualia of sorts. Makes perfect sense to me.

Perhaps that's not evidence to you, but I don't know what more you'd expect. I think you conflate consciousness with human consciousness, so you only perceive consciousness when it's similar enough to your own. If I may indulge in a silly (but, I think, accurate) comparison, it's a bit like a shoemaker conflating craftsmanship with shoemaking, and then saying there is no evidence that a boatsman is a craftsman, because they've never seen them make shoes.


Nonetheless, there is no reason to expect a description to be able to trigger the experience even if we knew exactly why it was triggered by sensory input.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: