Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

I would love to see a survey of the group of people who find Penrose compelling.

Anecdotally, I find the people drawn to his view are largely physicists. And the people who scorn him are largely "cognitive scientists". (I am closer to the second group than the former).

What you find to be "reasonable mechanisms", I find to be completely unreasonable. No, I don't think entanglement is a "reasonable mechanism" for "the fact that a bunch of disconnected neurons produce one mind."

I don't even know why one would think that quantum entanglement is even vaguely relevant to the "problem", and I think it has something to do with a misunderstanding of the problem.

Do you think that there is a problem of mind in the form of "a bunch of disconnected neurons produce one mind"?



I don't agree with Penrose's hypothesis, I just happen to agree that strange quantum behavior is related to the character of our consciousness. Other than that my views are completely different.

There is definitely a "binding" problem. Beyond the question of why the state of many disconnected neurons (or to go one level down, molecules) are subjectively experienced as a single consciousness, you also have the issue of why other neurons nearby are not subjectively experienced. After all, your consciousness doesn't even extend to your entire brain.


You don't actually give any reasons.

Why do you find it unreasonable that entanglement would be involved with integrating signals from disparate regions or several neurons?

The problem is how we integrate the signals from many regions in to whatever is generating the single perception (or perhaps several parallel perceptions -- I don't know that there aren't other experiences coincident with mine, just that I have access to one of them).

Unless you're postulating that a single neuron at a time is responsible for my subjective experience, then you do need to explain how several neurons are generating a single signal.

My experience with cognitive scientists is that they simply punt on the problem, completely failing to address how the signal is amalgamated in to a single stream of experience even as they talk about what regions contribute features of it.


Because that isn't the problem. It is not a correct or good statement of a problem.

Let me put it this way, suppose there was a mechanism, call it X, that explains to your satisfaction how many different neurons can "integrate the signals from many regions".

Does that explain to you why physical neurons create a subjective experience? Would you now consider the "problem of consciousness" to be solved?

Leaving "consciousness" aside, there is no issue at all of course. We have a straightforward computational theory of how different neurons can integrate their output and produce a computation.

The problem is that this idea of "integrating" signals is not clearly laid out. If it is not the computational problem, then what sort of problem is it?

Koch and Crick have also proposed "a mechanism" for something like the co-ordination of various neurons to explain visual awareness. A frequency at which all the neurons "cohere." (Koch is also a real genius, by the way. https://profiles.nlm.nih.gov/ps/access/SCBCFD.pdf )

Again, the problem with their proposal is that it is not really a "mechanism." Suppose all the neurons that recognize a face happen to be vibrating at 40 hz, Does that "explain" "consiousness?"

I lean towards the philosophers. Read one good critique by Fodor, and you realize that the questions are poorly formed.


It proposes a way by which those neurons can create subjective experience, yes. Namely, that the medium they're creating the signal in is fundamentally experiential and perturbations in it are fundamentally experiences of some variety. There are lots of philosophical reasons to think that this is the case for some aspect of reality (or perhaps fields in general, even).

This would give us a science of experience and allow us to categorize and create new experiential beings. That's a goal that many of us have.

Now, for various reasons, we might suppose a medium we're already aware of (or mediums in some combination -- as is the position of the materialist). Then the question becomes entirely about the integration mechanism, as that's the only piece of the puzzle we don't know.

So if you're a materialist, the integration of brain patterns in to a signal which corresponds to subjective experience is the question of creating new 'souls' (in the sense of beings who experience), and also of categorizing what exactly has a 'soul' (in the sense of inner experience).

The problem is how the computational structure which produces our behavior couples to our experience of it, which is a consistent, evolving perturbation in something, even if just in the sense of being a quasiparticle formed of the constituent parts interacting. (Though, likely, we're missing parts of the computational story -- I don't think most serious scientists would argue that.)

I think you just don't like the question, but it's pretty clearly formed, at least as far as big research directions ever are.


I am sorry but you lose me here.

"It proposes a way by which those neurons can create subjective experience, yes"

I just don't see it.

Suppose you somehow show me that various neurons contain particles that are entangled. Let's say I believe you.

So now how do they create subjective experience?

I see blue because my neurons are entangled? I see a flower because my neurons are entangled?

What is "the medium that they are creating"? How are you even using these words.

If I knew that my neurons have entangled particles, I would know no more about consciousness than I do now.


Particles don't create a medium, they're constructed out of perturbations in a medium, eg fields. Your paraphrase in quotes isn't anything like what I actually said in my post -- you dropped key words from it.

What I actually said, and you misunderstood: "the medium they're creating the signal in". Your paraphrase is so far from that, I have trouble even addressing the confusion. It's particularly egregious that you omitted key words when you're clearly using them to make passive aggressive comments about my understanding.

The answer to how this addresses the question is that the medium is experiential -- anything that creates effects in it is 'experiencing'. Experiencing seeing blue corresponds to a particular pattern of activity in the medium. (Though, there is no default 'experience of seeing blue' -- it's possible that we experience different ones, different animals have different ones, etc. This relates back to the computational structure.)

How entanglement is involved is that there's a single 'unified' experience when we know that different portions are generated in different brain regions, but don't (obviously) converge the information to a point (eg, it's always spread at least in a cluster of neurons), which suggests that we're talking about something non-localized in the medium, ie, not a point object.

The source of a lot of non-locality is through entanglement, so it's reasonable to conjecture that it's involved here, as well -- though, of course, it could be a different mechanism.

If the pieces of your brain which are generating the perturbation in the experiential medium(s) are entangled, it could provide a model of how it's carrying a non-localized piece of information -- your experience.


>The answer to how this addresses the question is that the medium is experiential -- anything that creates effects in it is 'experiencing'. Experiencing seeing blue corresponds to a particular pattern of activity in the medium.

But if your theory is that consciousness arises as a result of particular patterns in a medium, couldn't that medium equally well be "the universe"? Why do you see it as necessary for the medium to be entangled? Why couldn't consciousness just be the result of patterns in non-localised particles?


The medium isn't entangled, particles in it are (with each other), creating a larger non-localized object out of the smaller, (more) localized particles.

Entanglement is a mechanism by which a single, non-localized value can be carried by pieces (none of which themselves carry that value).

And hence the materialist proposal here is precisely what you propose: that the universe is fundamentally experiential (or at least parts of it that brains interact with), and the single, non-localized experience is created and carried via entanglement of the constituent "particles", which are themselves impacted by the regional computation of the brain (eg, by knotting with the "main" knot carrying the experience, in analogy to a TQC), thus allowing the disparate regions to contribute to a single experience with characteristics picked up by regional contributions.

Proposing that it's not entanglement means proposing a new phyical process by which that information can contribute to a single, non-local value, which people researching the brain haven't even taken a stab at.

Which is why I find their proposal that it's not entanglement to be strange -- they seem to blithely ignore the physics/informatiom theory implications of that.


>Proposing that it's not entanglement means proposing a new phyical process by which that information can contribute to a single, non-local value, which people researching the brain haven't even taken a stab at.

But you're already proposing something so fundamentally different to anything that exists, namely consciousness arising from patterns. There is no reasonable reference point to make any assumptions about such a thing.

>they seem to blithely ignore the physics/informatiom theory implications of that.

We can't make physical information assumptions about something which as far as we know may not even be physical. There is no obvious space-time location that consciousness exists at such that it needs to gather information to that point.


> Proposing that it's not entanglement means proposing a new phyical process by which that information can contribute to a single, non-local value, which people researching the brain haven't even taken a stab at.

I've never seen a brain researcher claiming to have seen any non-locality. It would be the kind of thing everybody would be talking about. Mostly, people seem happy to accept many speed-of-light delays on something that is centimeters wide and counts time in dozens of milliseconds.

You want a way to keep coherence overall through the brain, but there's no reason to think that coherence is needed.


Except that researchers also haven't seen the information converging to a single point, which should be obvious if it's happening.

I agree that the computing is done with moving charges and chemical propagation (at substantially below the speed of light), but that doesn't account for how the information gets integrated in to a single signal for experience. But we also see that kind of structure in, eg, the proposal to build a TQC, where you build a computation in to non-local structures built out of entangled particles by moving around charges at substantially lower than the speed of light.

If it's not being amalgamated in to a point object (which doesn't seem to be the case -- there's no 'you' point in the brain), then it must be being amalgamated to a non-local object.


I did cut off your sentence, and it was inadvertent.

I am happy to rephrase my question as what is "the medium they're creating the signal in?"

You are clearly getting annoyed at this, and I don't see us making much progress. I find this "experiential medium" even more confusing, and I don't see the connection between entanglement and "unified" experience.

But, hopefully others will make up their own minds based on our words.




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

Search: