It’s pretty nice to see Kurzgesagt finally continuing its “Big Questions of Life and the Universe” series. I shared the first part on consciousness over a year go. The series is funded by the Templeton Foundation, which I know many people have issues with, but so far the content has been reasonably scientific.
As they did in the first video, Kurzgesagt makes a distinction between intelligence and consciousness. But if you go back and watch the consciousness video, you’ll see a lot of overlap. I personally think the distinction is an illusion, a holdover from Cartesian dualism. When we let go of those notions, it becomes easy to see consciousness as a type of intelligence, specifically related to the spatiotemporal and action selection needs of a physical system trying to achieve its goals.
This video actually showed up when I was contemplating whether to comment on an article in Nature arguing that artificial general intelligence (AGI) will never be achieved, because AI is “not in the world”, by which the author appears to mean it’s not embodied.
The article is typical of a genre that identifies a current difficulty or obstacle, declares it unsolvable in principle, and writes off the entire endeavor as misguided. You’d think the ignominious history of people declaring it impossible for technology to reproduce some natural process would dissuade these types of proclamations, but people continue to do it.
Anyway, the difficulty is that the most primal processing in current technological systems is symbolic. But humans begin as physical beings in the world, building a world model, and then add symbolic thought on top of that. For us, all symbols ultimately reduce to some form of primal sensory or motor experience. When we say we understand something, we mean we can map it back to that primal world model. The argument is that as long as technological systems begin with the symbols, they won’t have general intelligence.
The lack of a world model is a real issue, and understood by many AI researchers. But it’s worth noting that we’re already starting to see systems with incipient world models. It’s what a self driving car uses to avoid pedestrians and other cars, or other semi-autonomous robots such as Mars rovers to avoid obstacles. These systems, compared to animal nervous systems, are still very primitive. They’re probably equivalent in spatiotemporal intelligence to early Cambrian animals (or perhaps late preCambrian ones).
A key question is whether a world model can only be built from the same type of immediate physicality that humans begin with. If we want AI systems to understand the world the way we do, it may be required. Or we may need to use it as an intermediate stepping stone until we understand well enough how this works to start from an alternative foundation. World models built from an alternate foundation may be unimaginably alien, but they may also able insight past many species level blindspots that humans contend with.
Only time will tell. But assuming it’s impossible, even in principle, strikes me as spectacularly short sighted.
Unless of course I’m missing something?
SelfAwarePatterns 
Re “When we let go of those notions, it becomes easy to see consciousness as a type of intelligence …” Actual my intuition is the reverse, that intelligence is a “feature” (or maybe a “bug”) of consciousness.
I wonder if that is really why the world goes around? Hmmm.
LikeLiked by 2 people
I guess it depends on how narrow or broad we define each of those terms: consciousness and intelligence. The video provides examples of intelligence where I don’t think consciousness is present. But if you define intelligence as only, say, symbolic thought, then you could see intelligence as a type of consciousness. And of course a panpsychist will see it that way no matter how broad of a definition of intelligence we use.
But I think a functionalist conception of consciousness is a type of intelligence.
LikeLike
I share your perspective Steve. If it’s true that we humans happen to be evolved creatures, then it should also be productive to say that there must have been a long string of characteristics which emerged at different times to create the human. Could it be said that sentience/ consciousness can exist in something that’s not “intelligent”? Given the premise of naturalism, I’m quite sure of that. Evolution must have added sentience to non-conscious organisms, and then progressively added what may now be referred to as “intelligence” given this subjective platform from which to work.
Mike,
I’m going to agree with you that the embodied cognition movement takes a bad line when it says that computers are “not in the world”. I believe that their message (which I do broadly share), needs to be refocused into a more tractable platform. Computers are of course no less “in the world” than my own body happens to be. Thus with the proper computer mechanisms (rather than generic information alone), this movement should formally state that of course a given humanly fabricated computer could, if built properly, function “intelligently”. Thus as their movement improves, yours should be robbed of an important scapegoat. Furthermore you should then be put in the position of defending the idea that qualia, unlike anything else understood in science, exists as the processing of generic information without associated instantiation mechanisms. Here instead of the strange “embodiment void” accusation that our side makes, your side should need to defend itself from the charge that it flouts nothing short of causality.
It was good for me to go through that Ragnar Fjelland paper in Nature. Though these are indeed my people, in order to get to the next level it seems to me that they’ll also need some guidance. I believe that they’ll need to demonstrate what it is that informationists propose specifically, or the very thing which my quite plain “thumb pain” thought experiment seems to.
LikeLiked by 1 person
Eric,
I think the embodied cognition point is a valid and important one. Brains are intimately entangled and inseparable from body operations. But some oversell and overdraw from that point to wedge in ideological positions, as Fjelland does. As you note, there’s no reason something like that relationship can’t be reproduced in technology.
I’d go even further and say there’s no reason it couldn’t be reproduced with a virtual body, although I suspect this will be where you part from me.
Don’t know if you watched the video, but did you notice it actually included a thumb pain example? (I thought of you when I saw it.)
LikeLiked by 1 person
Mike,
We certainly do part ways regarding your claim that a virtual body could theoretically feel what you and I do when our thumbs get whacked. It seems to me that just as the brain could only have qualia producing mechanisms in a natural world, “a virtual body” would by definition not have such mechanisms. So beyond my paper to paper thought experiment, this would be another illustration of what I consider “not of this world”.
It seems to me that I’ve presented plenty of reason to suggest that in a causal world, qualia should only exist by means of associated mechanisms. There’s nothing funky about something existing like that — it’s standard. But why would you say that there’s also nothing funky about something existing as generic information alone? Note that you can’t just leave this as open as “Because of neuroscience”. That’s not specific enough. If you were going to convince someone to disregard my argument and go with yours, what would you say? Do you know of anything else which is considered to exist as generic information beyond associated mechanisms?
LikeLiked by 1 person
Eric,
This point seems to be an obsession for you. We’ve been over this many times. I’m not sure how productive it will be to do it again. If we do, I’m wondering if there’s anyway we can be more productive about it.
The data from neuroscience shows a system of selective and recurrent propagation of electrochemical signals, with some stochastic perturbation from chemical balances and various environmental factors, such as electrical fields, although the system uses redundancy (concurrent connections and repeated signalling) to minimize the effects of noise. There’s also the supporting glia infrastructure, which may have some limited participation in the signalling. That’s what we have to work with.
Now, we can engage in loose speculation and fit all kinds of things into the current gaps, such as electromagnetic or quantum consciousness, property dualism, and a whole range of other notions. But when we do that, we’re wedging in things that aren’t in the data, explicitly or implicitly.
How would I argue against your position? Well, to the extent I understand it, I think it’s positing dualism-lite, not the supernatural form from religious traditions, but a physical kind. It’s loose speculation to provide something like the ghost in the machine, which I see no evidence for.
On your final question, I honestly don’t know what you’re asking for. My past interpretations (multi-realizability, information processing with no I/O, some notion of non-physical processing) haven’t appeared to match what you mean. Would you mind elaborating?
LikeLiked by 1 person
Mike,
Yes I suppose that you could say I have an obsession with this. What it essentially boils down to is that it provides me with hope. If it’s true that the embodied cognition movement needs improvement (which we both just agreed with), and that my thumb pain thought experiment does add substantially to John Searle’s quite famous but failed efforts, then I’d absolutely love to find a way to get this perspective out for wide consideration. I can see why you haven’t considered our discussions on this very productive, though I most certainly have. To me they suggest that my proposal does have legs.
You’ll get no challenge from me regarding any of that. Well said! But this premise not only supports your “qualia as information processing alone” position, but my “qualia by means of mechanism based instantiation of processed information” position. Mine essentially adds an extra step to yours. Instead of information processing itself producing something (somewhat like a prayer would need to?), from my perspective the information processing must animate associated mechanisms (or not just a plan, but a plan put into action). I don’t know of a single case where something is thought to exist without associated instantiation mechanisms. So yes, my model is implicitly in the data given the premise of naturalism, while the informationism model seems to remain incomplete.
To elaborate on the example that I’d like you to provide to help support the idea that qualia wouldn’t be the exclusively understood element of reality which exists by means of processing alone, let’s begin with a computer that’s processing some information. Clearly such information might then get converted into electrical signals which animate the function of a loudspeaker, for example. We wouldn’t expect to hear associated sounds by means of the processed information alone! So that’s specifically what I’m asking. If you believe that qualia is something which exists by means of processed information without instantiation mechanisms, then can you think of anything that’s actually established to exist in such a way? Note that even heat and entropy have instantiation mechanisms.
(I’m not sure what you mean by “multi-realizability”. Next, “information processing with no I/O” seems to essentially just present my question rather than provide an answer to it. On “non-physical processing”, actually I consider all of reality to function “physically”, which is to say to function by means of causal dynamics. But perhaps you’re referring so something that you’d call non-physical and I’d call physical? In any case, if you do have a situation which you suspect would make the status quo qualia proposal non-unique, then I’d love to hear it.)
LikeLiked by 1 person
Eric,
Well, I’m glad one of us is getting value out of these discussions. For me, the last several iterations have felt increasingly redundant with the end inevitable. It seems like if we’re going to continue having them, there needs to be new information to discuss. Otherwise we need to just agree to disagree.
On adding an extra step, the question I always ask for these kinds of additions is, what’s motivating it? Is there something in the data that requires it, or at least implies it? If so, I’m very interested to know what it is. Or are you just adding it to produce a result you prefer to see? If the latter, then it seems like loose rather than rigorous speculation. Rigorous speculation is a much stronger candidate for reality. Loose speculation is a story wedged in the gaps in evidence.
Multi-realizability simply means the same functionality can be reproduced in another system with a different underlying physics. It’s generally true with software (with exceptions).
On information processing with no I/O, you yourself gave an example with
“let’s begin with a computer that’s processing some information. Clearly such information might then get converted into electrical signals which animate the function of a loudspeaker, for example”
Prior to getting converted to electrical signals (actually it would be electrical signals the whole time, but I think you meant the proper type of signal for a sound system), it’s just information processing. Yet it has causal effects.
On something existing without “associated instantiation mechanisms”, you later state, “ I consider all of reality to function “physically”, which is to say to function by means of causal dynamics”. So you agree that all information processing is physical, and requires “associated instantiation mechanisms”. To posit information processing without a physical mechanism is incoherent.
So your real beef doesn’t appear to be something existing without instantiation mechanisms, but it existing with instantiation mechanisms you deem to be insufficient. So, my question would be, what about it is insufficient? Why is it insufficient? What properties are necessary to make a mechanism sufficient? What specifically about those properties adds that sufficiency?
LikeLiked by 1 person
Mike,
I can’t say that I’m a big fan of agreeing to disagree, or at least not among reasonable people who would rather be enlightened regarding any erroneous positions that they hold. I presume this of each of us. I generally agree to disagree with theists for example, since they often seem indoctrinated with a premise of “faith over reason”, but you’re nothing of the sort.
The situation here is that either the “informationism” position presumes a void in causal dynamics, as I believe, or it doesn’t, as you believe. It shouldn’t be new information which clarifies which of us is wrong, but rather a clear enough display of what it is that the erroneous position implies. And in the event that the erroneous party does happen to be too invested for such a lesson to take, well at least the other might use the experience for productive discussions with less invested people.
On the “extra step” (which I consider necessary rather than extra), what motivates it is the charge that nothing causal is produced by means of computer processing without associated instantiation mechanisms. That I prefer this step and you don’t, matters not in the end. In order to not be cast in the wrong direction, one of us is in need of correction.
On the concept of multi-realizability, thanks for the clarification. Right. Not only are there all sorts of physics based mechanisms from which to compress air into a tank, but to build the Pac-man video game. None of this however displays something that exists as information processing without associated actualization mechanisms that I know of.
On my example of processed information having the causal effect of animating a speaker, sure. It’s not my point however that processed information has no causal effects. Instead it’s that it won’t have the intended causal effects unless it animates appropriate mechanisms. So the processing of information doesn’t give us sound waves, that is until the speaker mechanism becomes animated. Similarly processed information could produce heat, though as a result of the machine which does the processing rather than the information itself. (This is simple to experimentally verify.) Conversely the informationism position is that qualia exist beyond any specific mechanisms, but instead concern the processing of information exclusively. That’s where I consider a supernatural line to be crossed.
On your final point, yes it is incoherent to posit information processing without physical instantiation. There is physical instantiation both regarding my brain, as well as a stack of information laden paper that’s converted into another. But note that this is instantiation of the processing rather than something else. The informationism position is that qualia is produced by information processing itself rather than any specific kind of instantiation. Again, that’s what I consider insufficient.
To hopefully help clarify what I’m saying here, the way around a supernatural premise would be to theorize the existence of a still experimentally undiscovered qualia producing mechanism. Given that neuron firing does produce electromagnetic radiation with a high degree of fidelity to that firing, to me this does seem like a promising (natural) potential medium which qualia might exist under.
LikeLiked by 1 person
Eric,
Agreeing to disagree is essentially acknowledging that we’ve repeatedly made our best arguments to each other, yet neither of us is convincing the other, at least not right now. It’s an understanding that we value the friendship enough to bracket this issue so we can have other interesting conversations.
As noted before, since I see information as causation, for me, saying that an information processing view presumes a void in causality is incoherent. Even if you disagree that information itself is causal, why wouldn’t the processing of it be so?
“On the “extra step” (which I consider necessary rather than extra), what motivates it is the charge that nothing causal is produced by means of computer processing without associated instantiation mechanisms.”
So are you saying the output of a logic gate which inputs to another logic gate isn’t a causal process?
“The informationism position is that qualia is produced by information processing itself rather than any specific kind of instantiation. Again, that’s what I consider insufficient.”
Once again, why? It’s trivial for me to find the information aspects of any qualitative experience. What about an experience implies something more? Until you provide an answer to this question, from my view, you’re simplifying repeating your bias as a mantra.
LikeLiked by 1 person
Mike,
It seems to me that our friendship is strong enough to thrive even given this particular disagreement. Perhaps this is somewhat because each of us practice the diplomacy of Mahatma Gandhi? Furthermore in practice I see no sign of waning interest on the matter. Perhaps it’s just more fun when we perceive ourselves to be making better points? Anyway, for now let’s open this up to your latest post.
Here we have thirteen modern consciousness theories which are being graded on the basis of nine forms of descriptive criteria. As far as I can tell, not one of these theories propose any form of qualia instantiation mechanisms beyond information processing alone. Correct me on that if you know of any exceptions. For example, Johnjoe McFadden’s proposal would provide such an exception, that is if included.
So how do you think things would go if I queried top proponents of each of these theories, in respect to my thumb pain thought experiment? Surely they’d each agree that information gets sent to my brain when my thumb gets whacked, resulting in me feeling the qualia that I do. But under their models, if it’s all just information that’s converted into other information, would any of them have the integrity to bite the bullet and agree with the implications of my experiment? Would any say, “Yes, if an enormous stack of paper depicting whacked thumb information were processed into another enormous stack of paper depicting the brain’s response [and hopefully they’d get into the specifics of a given position], then something in this paper to paper conversion would thus feel what you do when your thumb gets whacked.” I highly doubt that any would have such integrity. Instead I’d expect all sorts of squishy exceptions and academic double talk. But what do you think?
Actually I do agree that information itself can be causal. (In order to avoid the fallacy of begging the question, I don’t think that the is term should be used there.) Note that information is the stuff which is currently animating my computer screen. And the processing of information into new information can obviously be causal as well. When I press a key, this information is processed in a way that ends up altering my screen pixels. But if I were to disconnect the screen and then press the key, what would happen? Here the computer should still process that information, though I shouldn’t see an associated screen effect. Thus it should not be productive to say that it’s the processing itself which alters my screen, but rather the processing in respect to a thusly animated screen mechanism. I know of no information based effect without associated mechanism for that information to animate, and certainly including any logic gates. (Note that they even put the “gate” mechanism term right in the name!) It seems to me that the informationism position posits qualia as the only exception to this rule.
As you say, it’s trivial to find information aspects of any qualitative experience. We’re together on that. But for qualia to exist, naturalism itself implies that there must be associated mechanical instantiation of that information. In a causal world, information with no mechanisms to animate, should indeed go unrealized.
LikeLiked by 1 person
Eric,
“Correct me on that if you know of any exceptions.”
For anyone reading this after the fact, the post is: https://selfawarepatterns.com/2020/07/15/hard-criteria-for-theories-of-consciousness/
Actually, Orch-OR, IIT, RPT, TLT, and NMDA all posit something beyond straight information processing. Although none of them actually explain why that extra thing is needed. (IIT and RPT actually do make an attempt, but RPT’s is vague and IIT’s seems like obfuscated hand waving.) I would put McFadden’s theory in the same category, equating consciousness with some magical property or phenomenon, but not explaining why that phenomenon in particular grants consciousness.
The ones that fit your description are the computational ones: GWT, HOTT, PPT, ART, AST. Many of the proponents would likely tell you not to waster their time with philosophical thought experiments. (Baar and Deheane have expressed disdain for such exercises.) The others would probably point you to whatever response they had previously made to the Chinese Room, since yours is just that with an affect thrown in.
“In order to avoid the fallacy of begging the question, I don’t think that the is term should be used there.”
It’s not begging the question to make an assertion. It’s only begging the question if justification of that assertion assumes its truth.
On logic gates, if you accept them as a sufficient mechanism, then since any actual information processing will involve gating mechanisms, you accept information processing as sufficient. You’ll no doubt deny this, but then I think whatever concept of “information processing” you’re holding onto here is incoherent.
“But for qualia to exist, naturalism itself implies that there must be associated mechanical instantiation of that information.”
Since information itself can’t exist or be processed without mechanical instantiation, again, this statement is incoherent. If you say that the mechanism sufficient for information is insufficient for qualia, then once again I’ll ask, why? You keep throwing up all kinds of peripheral matters and evading the core question. Why do you think something other than information is needed?
LikeLiked by 1 person
Mike,
Thanks for the clarification about which of these theories are purely information based and which are not. Now that you mention it, IIT proponents (as I recall) essentially state that everything is conscious to the extent that its information becomes “integrated”, and in practice this only occurs through the function of living systems. So they’ve put together some sort of phi score to estimate how conscious something is on the basis of such life based information integration.
Regardless, surely IIT proponents would thus not consider it possible for information laden sheets of paper converted into another such set, to create something which feels what I do when my thumb gets whacked — life based function would instead be required. I have no idea why living stuff which “integrates information” should be considered conscious, but then to me all forms of panpsychism seem ridiculous. Any thoughts about what the other three such approaches would say regarding my thought experiment? I presume that they’d all deny that qualia would occur, though I doubt than any of their accounts are as simple as mine is.
It now occurs to me that it should be helpful for us to establish where we agree so that our disagreement can be reduced as far as possible. We each agree that there are mechanisms in the brain which process information (essentially centered around neuron firing based logic gates as I understand it). Furthermore we each agree that information on paper can algorithmically be converted into another set of information on paper through all sorts of devices. And for this let’s consider a vast hypothetical supercomputer which accepts such paper based information (or input), and processes it into another such stack (or output). So in a conceptual sense we each agree that whatever information processing that a human brain does, could also be done by such a hypothetical supercomputer as well. Correct?
The essential difference in our perspectives here as far as I can tell, is that you believe qualia exist as brain information processing in itself. Thus the very different mechanisms by which the same information on paper become converted into a corresponding output set of information on paper, theoretically would produce something that experiences what we do when our thumbs get whacked. So here’s my simple account of why something other than processed information should be needed in addition to this.
In a causal world I believe that qualia must exist as more than processed information alone, because we only know of situations where information animates various production mechanisms, such as loudspeakers for their sound, computer screens for their images, and even basic logic gates for their gating function. So while I see the informationist account as supernatural, since the secondary physics of mechanical instantiation goes missing for qualia production exclusively, you see it as natural given that you consider qualia to exist as the processed information in itself.
Regardless of which of us is wrong here, do you accept this account as a reduced explanation of our difference of opinion? Have I now clarified my (perhaps erroneous) objection to your position?
I consider there to be all sorts of problems with Searle’s famous Chinese room thought experiment. What I’d like to do is fix those problems so that the academic community might then be able to more effectively understand the nature of various consciousness proposals. If your perspective does happen to be the correct one, then surely it will be validated in the end regardless of my clarifications. Good science requires good competition, and I’m afraid that my side has not preformed as well as it might so far.
LikeLiked by 1 person
Eric,
All the ones I listed as positing more than information processing, such as IIT, would not consider the scenario you describe as conscious. In the case of IIT or RPT, you could arrange things so that they would, such as having billions of people, each passing paper between each other, in looping patterns (recurrent – RPT) and integrating information (IIT). I don’t know enough about the others to comment further, although I think Orch-OR requires quantum superpositions.
Along those lines, IIT doesn’t restrict consciousness to only living things. You can arrange matter in such a way that it has arbitrarily high phi scores. Apparently Tononi bites the bullet and says these are conscious.
“So in a conceptual sense we each agree that whatever information processing that a human brain does, could also be done by such a hypothetical supercomputer as well. Correct?”
Yes, in principle.
“Regardless of which of us is wrong here, do you accept this account as a reduced explanation of our difference of opinion?”
I think so, although I have a question. You use the example of a speaker producing sound. So it’s producing something that is then outside of the system. I think I know the answer to this, but I’d like your explicit clarification on whether I have it right. Are you saying that qualia exist as something output by brain processing, that then exists separate and apart from the brain? It seems like with the EM and second computer stuff, you are.
If so, then I think we have two disagreements. First, I see qualia as information processing. Second, I see it as completely happening in the brain, specifically the neural circuitry. There is a constant feedback loop between the brain and body, which does heavily factor into this, but what reaches the brain is information. (Although the brain’s ability to function can obviously be affected by nutrient availability, fatigue, hormones, and other environmental factors.)
I don’t know that you’re going to get much mileage tinkering with the Chinese room. It’s a rhetorical device that works for people with certain biases. For people without those biases, it’s, at best, irrelevant.
LikeLiked by 1 person
Well done Mike! I see this as a model of effective discussion regarding a given disagreement. Politics will always have the potential to rear its ugly head, and that’s given that we’re all self interested products of our circumstances. But it’s remarkable to me how few seem to grasp even here that the strategy of the great Gandhi not only fosters effective discussion, but effective political maneuvering. What was that Ben Kingsley line regarding his portrayed path of civil disobedience against the British authorities? Something like “Yes, so that you yourselves will see the wisdom of walking out of India”. I hold such beacons dear.
(On the topic of sensible people, I see from your twitter feed (https://www.nytimes.com/2020/07/15/us/steven-pinker-harvard.html) that Steven Pinker has gotten into some trouble with the prevailing social justice warrior crowd. It’s nice that occasionally some of us are able to say sensible things without being squashed by such forces. I know a Lily white privileged young woman who wasn’t satisfied with simply posting an off the shelf “Black Lives Matter” poster on the lawn of her high dollar gentrificated urban home. Instead she seems to have scored extra clique points by making a custom sign which added the “Trans” term as well. And do many of the people in the not yet generificated surrounding neighborhoods also refer to her as “a Karen”? I’m sure of it.)
It’s interesting to me that Tononi has had to bite a bullet regarding some queer implications of his extremely popular IIT. Note that such an admission could only have occurred by means of a thought experiment. So it seems to me that without thought experiments we wouldn’t be able to display what it is that a given consciousness theory effectively entails, as well as ideas in general. This doesn’t mean that all thought experiments happen to be helpful, but rather only to the extent that they do provide effective demonstrations of what a given model suggests.
Regardless, imagine billions of people passing notes between them in looping or integrating patterns, which thus create something that experiences what you and I know of as “thumb pain”… Thus we can say that ideas with these sorts of implications pass muster regarding consciousness exploration today. Without a community of respected professionals armed with effective principles of metaphysics, epistemology, and axiology, relief should be slow.
You and I disagree with Searle’s Chinese room in many corresponding ways. Thus for my own version (which you did help me create) people aren’t able to make those specific objections. So from your perspective, how could my version not more effective than Searle’s?
The most important improvement, I think, is that instead of depending upon people to grasp what Searle meant by “understanding a language”, (which opponents are thus able to effectively bastardize into a non-qualia based idea), we should all be quite certain about what it feels like for our thumbs to get whacked. So here I present a far less difficult idea to misinterpret than Searle. Furthermore instead of depending upon hand written notes, I leave the processing to an arbitrarily fast computer. This effectively takes away what I’ve noticed to be your favorite Chinese room objection. I also clarify that I’m as strong a computationalist as they come — it’s instead “informationism” which my argument suggests otherworldly. And do I thus consider consciousness to be “life based”? No, I consider it “physics based”. If the processed stack of paper were then fed into a machine armed with appropriate qualia instantiation mechanisms, then as I see it something should indeed experience what I do when my thumb gets whacked.
Anyway there are some heavy discussions which I’d like to have with the modern embodied cognition community. I believe that we could help each other out quite a bit.
On your question regarding energy waves, whether vibrations, electromagnetic radiation, or anything else, I do have some clarifications to make. First I should disclaim that as a naturalist the “systems” that you’re referring to must be considered merely epistemic. The only “outside” in an ontological capacity, would be a supernatural realm.
That said, I don’t believe that I do run afoul of your second concern. Just as you consider there to be feedback loops between brain and body (as I do), this may also be said of brain to the consciousness medium. In fact I’ve had a psychology based model denoting the nature of this relationship at my disposal for over a decade. (I smiled when I saw that one of the criteria for assessing a consciousness model from your noted post, was how it deals with the conscious/ non-conscious relationship. I doubt anyone provides a more distinct separation on this than I do.)
Back when we met nearly four years ago, if you recall I had no idea what substrate consciousness might exist under. I’d tell you that the entire brain may effectively be analogized as a non-conscious computer, and that under the proper conditions this machine produces the consciousness dynamic by which we experience existence. I’d called consciousness a minuscule teleological form of computer which effectively conditions the non-conscious brain that produces it. Here you’d ask (I presume smiling) where this other computer might be if it’s not itself “brain”? I’d say that I didn’t know, but clearly the brain itself was not conscious, and yet consciousness does exist.
It was merely this past December over at James Cross’ blog that I realized the potential for electromagnetic radiation to exist as my long theorized consciousness instantiation mechanism. Think about this for a second. If you weren’t so utterly convinced that an enormous stack of information laden paper which is somehow quickly converted into another such stack, could thus produce something that experiences “thumb pain”, which is to say if you saw some potential that qualia might require a given kind of substrate in order for it to exist, then what other brain based substrate might suffice? We know that neuron firing produces associated radiation. What other product of the brain could harbor a sufficient level of information fidelity?
LikeLiked by 1 person
Eric,
I actually have no issue with people supporting Black Lives Matter or LGBTQ causes. Quite the opposite. I’m happy that our society seems to finally be dealing with the way we’ve treated these people. But the attack on Pinker wasn’t that. Nor was it merely an attack on his ideas and opinions, which I think is totally fair game. It was an attack on his professional credentials, his livelihood, because of his ideas and opinions.
Now, if someone is advocating for violence, or expressing opinions that clearly show they’re not competent to do their job, then such an attack might be warranted. But Pinker merely expressed views they disagreed with. And for that, they wanted to have him thrown out of a professional organization. That’s not free speech. It’s speech suppression. Intolerant thought policing is not right regardless of whether it comes from the right or left.
I’ve stated before that I do think thought experiments can be useful for clarifying our intuitions, which is what Aaronson’s thought experiment with the expansion boards did. He took the mathematics of IIT and applied it to a system with arbitrarily high phi to demonstrate that it makes predictions that violate our intuitions. It doesn’t prove the expansion board is not conscious, merely that IIT predicts a system to be conscious that doesn’t meet our intuitions of consciousness. And as Aaronson said, when it comes to what is conscious, our intuitions are all we have.
But just as the Chinese room can clarify our intuitions, it can’t tell us which intuitions are right. Sorry, but your modification can’t overcome that. Someone whose intuition is that a computer can’t have qualia will nod in agreement. Someone who dismisses that intuition will still find the argument unconvincing.
When you talk about substrates outside of the brain, I take that to be confirmation that you hold qualia to exist outside of the brain. So I do think we disagree on those two points: whether qualia is more than information processing, and whether it happens outside the brain. I think the answer to both is no.
I’m open to being convinced otherwise, but it will take evidence (widely accepted and reproducible evidence), or at least logic showing it to be the most parsimonious explanation. Given that this is an area where people have very badly wanted to find such evidence for a long time and failed, I think we have to be on extra guard when someone claims to have found it, particularly when they seem unable to convince a significant portion of the neuroscience profession.
LikeLiked by 1 person
Mike,
My point was not that we should oppose these social causes. I do support fixing injustices of all stripes. And it seems to me that massive progress has been made in recent years, or at least regarding LGBTQ. I’m sure you remember how different things were ten, twenty, and thirty years ago versus now in that regard. (Expect to get an earful for mentioning “progress” to one of the modern justice warriors.)
Beyond the censorship, what I don’t support is that people now seem cowed into not thinking for themselves. It’s as if we’re raising an entire generation of tools who seek to one up each other through hollow displays of progressiveness. Of course the backlash to that helped give us Trump.
I agree that thought experiment intuitions don’t give us truth, though there is something that should be more difficult to dispute. As things begin firming up in our mental and behavioral sciences, when psychologists use the “consciousness” term in general they should progressively mean some reasonably understood specific idea. My guess would be “qualia based function”, but we’ll see. And regardless of what shape this term ends up taking, there should eventually be ways to empirically measure it.
Just a second. I also happen to be a person who believes that humanly fabricated computers could produce qualia by means of associated physics based instantiation. I even propose that EM waves might be responsible for such brain function, or physics which is well in the scope of human technology. So with this statement you’re not actually challenging what I’m saying. Though perhaps an oversight, this also seems a bit strawmany.
Regarding EM fields existing outside the brain, perhaps you misspoke. Because neurons fire inside the brain, this radiation must indeed be found there. Another oversight?
Though I do consider the various interconnected models that I’ve developed to be quite parsimonious, my thought experiment doesn’t get into any of that. Instead it suggests “informationism” to be overly parsimonious. As I see it such models skip the mechanical instantiation dynamic by which all known aspects of reality are understood to occur. But even if otherworldly, parsimony would still be parsimony. As long as we don’t call it natural, I’m fine saying that informationists present more parsimonious models than I do.
Isn’t IIT one of the two most popular consciousness theories in science today? If that’s at all the case then it seems to me that informationism isn’t getting a completely free pass yet. And that was when it merely had to fight the Chinese room. If the embodied cognition people were to make some modifications then they might put up a better fight. I’ll do what I can, though speaking with an opponent who’s as strong and convicted as you are, does seem to be getting me into shape on this!
LikeLiked by 1 person
Eric,
The problem with defining “the consciousness term” is that any objective definition is theory laden, and therefore controversial. An IITer will say that consciousness is high phi. A GWTer, global broadcasting. A PPTer, prediction. An ASTer an attention schema. I suspect you reject all of these.
People often try for theory neutrality by giving subjective or phenomenological definitions. Unfortunately, these usually amount to little more than synonyms such as “subjective experience” or “phenomenal experience”. For yours, “qualia based function”, you now have to define “qualia”, which is usually presented as instances of subjective experience, which just puts you back in the same boat.
“I also happen to be a person who believes that humanly fabricated computers could produce qualia by means of associated physics based instantiation.”
Sorry Eric, but that’s not computationalism. It might be computationalism plus something else, but it’s not straight computationalism. Most people think the brain does at least some computation. They just want to add something else. Of course, no one can agree what that something else might be, not to mention find any evidence for it.
Neurons fire in the brain, but you’re positing that qualia are in the electromagnetic field, which permeates but also goes beyond the brain. There is such a field of course. It’s used in EEG and MEG scanning technologies. But there’s no evidence for the kind of stuff EM theorists posit happening with it, at least no widely accepted reproducible evidence.
The only reason something would be “too parsimonious” is if it doesn’t address all the evidence. What evidence does straight computationalism fail to address? (Note: saying “consciousness” or “qualia” would just be begging the question.)
IIT is more popular outside of consciousness research than in it. A lot of people who aren’t very familiar with it are impressed by the fact that it’s mathematical. I’ve linked to this survey a few times: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6230957/ (Scroll down to figure 3.)
LikeLiked by 1 person
Mike,
On defining consciousness, I wasn’t actually referring to our situation today with its IITers and on and on. Surely you believe that thing’s won’t always be as pathetic in our mental and behavioral sciences as they are currently? Well given such progress, in some number of years I’d expect that when a psychologist uses the “consciousness” term, that there will be a reasonably accepted definition for it as well. How could “hardish” scientists continue using a term without giving it a specific functional meaning as well? Here I’d also expect some experimental progress measuring whatever it is that these progressed scientists mean by the term.
Consider this potential. It could be that in ten years there will be algorithms from which to interpret the EM waves associated with neuron firing, which also correlate highly with what’s reported as “feeling bad”, and/or “feeling good” in people. (Yes, in opposition with Lisa Feldman Barrett’s research and theory of constructed emotion.) And let’s say that we not only find these patterns in the human, but for all sorts of creatures that we presume we are punishing or rewarding. Here we’d have nothing short of an effective physics based definition for qualia / consciousness — the observed forms of electromagnetic radiation.
Then let’s say that we not only become able to reproduce such waves, but somehow provide the theorized experiencer (which would presumably exist as those waves themselves) with a way to choose to diminish or amplify such theoretical experiences. Then let’s say that with enough research scientists deduce that it must indeed be the proposed experiencing entity which “chooses” to feel more good and less bad through these mechanisms that we provide it. So here it’s decided that the choices aren’t made by some errant product of the system, but rather an actual teleological agent. So by knowingly creating qualia, we’d effectively be creating a non-functional conscious entity, or the very thing which evolution apparently used to add a functional teleological dynamic to many of its creations.
What would this effectively do? Well the conventional wisdom is that science progresses one funeral at a time, so perhaps many would go to their graves asserting that qualia requires no instantiation mechanisms regardless of such evidence. If fruitful however then this should mark a substantial paradigm shift nonetheless.
While the general public should then want their own sentient machines as servants and friends, (and I suspect there’d quickly be laws making it illegal to build any machine which is ever permitted to feel negative), I suspect that the main effect would be to foster progress in fields like psychology, psychiatry, sociology, and so on. Though we may never be able to practically build very much in the popular sci-fi genre given associated engineering challenges, with this advancement we should finally be able develop effective general models regarding our nature. To me that would be a feat far exceeding any specific associated technology. Here we’d finally begin exploring ourselves as a physicists explore physics, which is to say, amorally.
One of the greatest blunders that John Searle and others on his side ever made, I think, was to not challenge their opposition when it took up the “computationalists” label without them. This not only should have placed what I instead call “informationism” under the respectable position by which brains function through computational dynamics (which should be obviously true to all but a fringe group), but should have effectively demoted their group to that very fringe. Huge mistake! Instead they should have also called themselves computationalists, and referred to their opposition by means of a far more appropriate name, such as “softwarists”, or “informationists”. I’d like to help correct this mistake.
Here’s a quick quote from Wikipedia which doesn’t to me suggest that the “computationalism” term must be reserved exclusively for people who believe that qualia exist as information processing alone:
In philosophy of mind, the computational theory of mind (CTM), also known as computationalism, is a family of views that hold that the human mind is an information processing system and that cognition and consciousness together are a form of computation. Warren McCulloch and Walter Pitts (1943) were the first to suggest that neural activity is computational. They argued that neural computations explain cognition. The theory was proposed in its modern form by Hilary Putnam in 1967, and developed by his PhD student, philosopher and cognitive scientist Jerry Fodor in the 1960s, 1970s and 1980s. Despite being vigorously disputed in analytic philosophy in the 1990s due to work by Putnam himself, John Searle, and others, the view is common in modern cognitive psychology and is presumed by many theorists of evolutionary psychology. In the 2000s and 2010s the view has resurfaced in analytic philosophy (Scheutz 2003, Edelman 2008).
The reason that your explanation regarding the creation of qualia is more parsimonious than mine, is because it’s more simple. Here qualia exists by taking one set of information and turning it into another — no need to use that second set of information to animate any dedicated qualia producing mechanisms. But parsimony also isn’t the end goal. For example, the most parsimonious answer that I know of for anything would be something like “Because God did it”. Though I can’t claim with certainty that it’s wrong, this does conflict with my own brand of metaphysics. Informationism, though more parsimonious than my own presumption regarding the creation of qualia (given that instantiation mechanisms go missing), nevertheless lies in opposition with naturalism.
LikeLike
Eric,
Well, I think if we’re being scientific, then we apportion our beliefs according to the evidence, and to the theories with the fewest assumptions that make accurate predictions. True, science can be said to progress one funeral at a time, but that refers to major figures ignoring what the evidence, and lack of evidence, currently show. How much of your reply cites evidence or successful prediction, and how much of it is rationalizing why we should ignore the current balance of evidence?
On pain, don’t know if you follow Richard Brown’s Consciousness Live podcast, but he recently did an interview with Jennifer Corns, who studies pain. I’m thinking about reading her book, but it’s expensive.
https://onemorebrown.com/2020/05/23/jennifer-corns-live/
LikeLiked by 1 person
Mike,
I’m not ignoring any evidence that I know of. But if you think that I am, then what evidence might that be?
The only quite good evidence against EM consciousness theory that I know of, would be to demonstrate that qualia does exist without such fields. My paper to paper scenario by which you think thumb pain can exist, would do it if validated for example. No EM fields there.
Last time I presented a scenario by which my model would be supported. This doesn’t seem nearly as fantastic to me My point however is that there’s nothing supernatural about proposing something which exists by means of informationally animated mechanisms. That’s standard. As for information itself existing as some output effect, and yet without the benefit of any mechanical instantiation, that would be interesting. I guess if you could come up with such an example in any capacity at all then this would help support your position.
I think you’ve suggested Richard Brown’s podcast to me before, and perhaps somewhat because he’s of the opinion that consciousness is “life based”. (Not that I go that way myself.) Anyway if it gets very deep into the neuroscience of pain, then Jennifer Corns’ book should be right up your alley.
Regarding the discussed position, I’ll generally disagree. As I understand it she advocates eliminating the “pain” term for scientific discussion in favor of using other terms, such as describing associated brain function. Why? Because apparently the mechanisms associated with “pain” as the term is commonly used are too complex and varied for effective “type” classification. (She has no problem with folk uses for the term though, which is good.) I see that in your recent post you mentioned Pascal Boyer, who says that we shouldn’t consider there to be any religions, and for the same reason — a common theme doesn’t seem to exist, or at least for technical purposes.
I like how Corns uses the “useful” criteria for the terms we use, which corresponds with my first principle of epistemology. There’s more to that principle however which puts me in opposition.
I believe that the theorist must be given the freedom to directly define speculative terms, such as “consciousness”, “religion”, “pain”, and so on, however they like. My EP1, if generally accepted, would obligate the reader to grant those definitions in the quest to assess whether or not useful ideas are being presented. I have no doubt that scientists are able to use the “pain” term, both in folk and analytical capacities, in productive ways.
For example, for certain surgeries anesthesia is provided given what the patient would otherwise experience. This is to say “treating the pain”. It should generally be helpful for scientists to have such a term to serve as reason for anesthesia, and I think a great deal more. And as scientists learn more about the properties which make body damage uncomfortable for us, the term should become more specific both in science as well as for folk uses.
LikeLiked by 1 person
Eric,
I think you’re ignoring mountains of evidence acquired through over a century of research on how neurons and synapses work, and the fact that evidence for communication via electromagnetic fields is scant and controversial, despite generations of people earnestly trying to find it, or read something into what is there. Maybe that will change, but until it does, to the extent there is an effect, it seems to be, at best, perturbational, participating in the stochasity of neural processing rather than meaningful processing.
I did end up buying Corns’ book. I think it’ll be next up after I finish Sapiens and a sci-fi book I’m reading. Her point that there is no consistent objective mechanism that maps to our folk conception of pain actually matches my overall conclusions about consciousness itself. So I expect to find her book pretty interesting. But pain will always be with us as a subjective experience, just as subjectivity itself will be, even if there’s no single objective mapping for those things.
LikeLiked by 1 person
Mike,
Your logic on this doesn’t add up to me. How can you decide that a century of neuroscience regarding the function of neurons and synapses tells us that qualia doesn’t exist by means of EM radiation, when the two always exist together? If there were ever a situation where neurons fired without EM fields, then it would be possible for a century of study to eliminate the potential for EM fields to exist as a mechanism for qualia instantiation. As things stand however, this potential has not existed.
Furthermore the informationism proposal is unprecedented regarding accepted ideas in science. Here something is produced (or qualia), without the mechanical instantiation theorized to exist by means of all computer and non-computer based function. Ultimately it must be faith which leads you to believe that qualia exist as information processing alone, since you haven’t been able to provide a second example of such function. Regardless, this is the line of reasoning which I think Searle and the embodied cognition people need to explore next. I’d like to help improve their game.
I’m looking forward to your review of Corns’ book. On “pain” and “consciousness” however, I guess I’m more optimistic than either of you. First I believe that there will be progress in epistemology which validates these still speculative terms for the purposes of science. Second I believe that mechanical explanations will eventually be found for qualia, and thus terms like “pain” and “consciousness” will end up being defined under this platform. Just as Newton made huge strides against substance dualism, this achievement should as well. “Faith”, however, should always be present for science to challenge by means of “reason”.
LikeLiked by 1 person
Eric,
On logic, your reasoning seems equivalent to a radio astronomer asserting that because there’s so much radio activity out there, there must be communications from aliens in it. In both cases, that’s just begging the question. You need additional evidence to justify your claim. The evidence you’re citing seriously under-determines that claim.
On qualia, I think the version you’re insisting must be explained, a version somehow generated but separate from the neural signalling in the brain, doesn’t exist. It’s naturalistic dualism. You see me failing to account for that naturalistic dualism, and assume the only other alternative is supernatural dualism. But I think the reality is no dualism at all. Until you can provide some evidence for that version of qualia, I see no reason to entertain more than what we do have evidence for, particularly when the proposition seems like just an attempt to save appearances for folk notions of consciousness.
I actually don’t see Corns’ or my view in any pessimistic or optimistic sense, except perhaps in the optimistic sense that we are getting closer to understanding the actual reality. Given the history of science, the fact that our folk intuitions are being challenged seems like a sign of progress.
Of course, those folk intuitions will need to be accounted for, and can always be viewed in an emergent fashion. I’m an emergentist, not an eliminativist. Seeing pain or consciousness as emergent phenomena doesn’t mean they don’t exist, only that they are complex, with an objective reality not readily apparent from our subjective impressions.
This simple principle, seems to be the most difficult one for many people to accept, but for understanding the mind, it’s essential. Those who can’t accept it seem doomed to be stuck with the Goff’s, Chalmers’, and Kastrup’s of the world, in a never ending search for magic.
LikeLiked by 1 person
I don’t know about that Mike. It wasn’t my claim that logic can be used to show that qualia exist by means of EM fields. As you imply, evidence would be needed for that sort of thing, not logic. Instead I was observing that because all known qualia are accompanied by the EM fields associated with neuron firing, that a century of research about how neurons and synapses work simply couldn’t have displayed this to be false. That wouldn’t be logical. Instead we’d need something more like a real experiment which demonstrates that qualia can also exist when no EM fields are present. My paper to paper to thumb pain scenario would do the trick, that is if something with qualia were displayed to exist that way.
I’m not sure why you have the impression that I’m pulling a “David Chalmers” here and supporting the so called “Naturalistic Dualism” position. I’m as strong a monist as you’ll ever meet. I make no exceptions whatsoever regarding qualia, quantum mechanics, or anything else. This is to say that if something isn’t ontologically caused to happen exactly as it does by means of the overarching causal system itself, then under my own brand of metaphysics it simply cannot happen. Though I do concede that I could be wrong about this, to the extent that I am, science itself becomes obsolete. This is to say that to the extent that my own naturalistic, monistic, physicalistic metaphysics fails, nothing would exist to grasp. Why try to understand that which has no potential to be understood? There’s a question for Chalmers and the rest.
With that said, I do have concerns about the purity of your own naturalism given the informationism premise. These positions seem inconsistent with each other. If there are causal dynamics associated with the creation of qualia, then in its most elemental state it must occur by means of mechanical dynamics of some kind. Note that all elements of computation, and all else accepted to exist in science, have this feature. So proposing that qualia exist differently from the rest, is unnatural. That’s why my paper to paper to qualia thought experiment is silly — no mechanical instantiation is proposed. But if I instead were to reference a system with the same mechanisms that the brain uses to create qualia, then it wouldn’t be silly. (And sure, that would be begging the question — in that case the machine would inherently have qualia. But that’s because it would have the same physics based mechanisms, not generic information alone.)
Regardless, this is an argument which clearly needs true academic consideration. John Searle’s “understanding a language” idea has gotten a lot of play. (I believe I saw a survey where he was something like the world’s 13th most cited living philosopher.) But he hasn’t gotten down into the weeds that my own arguments are able to.
LikeLiked by 1 person
Eric,
The number of propositions that have not been falsified is vast. The number of propositions that are implied by the data, even if not yet verified, that are probable candidates for reality, is much smaller. And of course, the number of propositions that have actual evidence for them is smaller yet. I think what you’re talking about is in the first category: loose speculation. In principle it remains possible, but the probability that it’s the right guess is very small. We’re talking Russell’s teapot territory.
I make no claim to any kind of pure naturalism, or any other philosophical ideology, particularly anyone’s narrow interpretation of it. I’ll use those types of philosophical ideologies to the extent they provide useful assumptions of what we’ll find in observations. But they’re only starting points. I’m not going to adhere to them in the face of contradictory evidence. Doing that turns them into a religion.
I’m not surprised Searle gets a lot of citations. Many people hate computationalism, and he made his name attacking it.
LikeLiked by 1 person
Mike,
Regarding what you currently consider “propositions that are implied by the data, even if not yet verified”, what if you were to eventually realize that some of your favorites, such as global workspace theory, violate the premise of naturalism? What if you were to decide that its truth would depend upon a failure in causal dynamics given that apparently information remains effect-less under a natural system when dedicated mechanisms for information implementation go missing? I’m asking which of the following is more important to you, informationism or naturalism?
If the latter, then it seems to me that you should currently be looking for reason to believe that causal dynamics exist by which information in itself, whether by means of neuron firing, inscribed on paper, or through any other medium, is able to create a common thing beyond the popular presumption of qualia. And perhaps you have thought about this, though please do if not.
I say this because if you and others are unable to come up with even one more example of generic information in itself that creates a common “something” by means of various information platforms alone (or no dedicated instantiation mechanisms for that information to animate), and yet remain committed to the informationism premise, then apparently this would be because informationism would trump naturalism for you. It’s a general question that I have for those who subscribe to the informationism premise.
LikeLiked by 1 person
Eric,
If I thought global workspace theories weren’t naturalistic, I’d probably dismiss them. But I can’t see that you’ve successfully demonstrated that yet.
Remember I see qualia as information processing and information itself as causation, or at least deeply entwined with it. Which means I’m already working with a fully causal account. It’s just not one you like. Until you can give me some reason why it’s inadequate, engaging with the rest seems unproductive.
LikeLiked by 1 person
Mike,
I agree with you that I haven’t been able to conclusively demonstrate that the informationism premise should be considered supernatural. What my arguments do seem to do however, is force informationists who hold naturalism dear, to make concessions in order to square the two. From here it’s up to general observers to decide if those concessions seem valid.
I can see how “causation” might be equated with “information”. Here we wouldn’t just say that causality mandates that a moving ball transfers a degree of its momentum to one that’s struck, but also by means of information. According to my EP1 you’re well within your rights to equate the two terms. Then it’s up to observers to decide whether or not a useful point is being made here as well.
The more tricky proposition for the naturalist, I think, would be to consider qualia to exist as certain examples of information processing. (Surely not all all such examples, but some.) Notice that here there’d be no question whether or not certain information on paper that’s converted into another such set, would create something that feels what you do when your thumb gets whacked. This conclusion would be mandated by the provided premise that qualia exist by means of certain examples of information processing.
It seems to me that given my arguments, informationists who aren’t prepared to beg that question might often look for alternative conceptual solutions regarding the creation of qualia. That’s the path that I offer. I believe that some would then wonder if there aren’t any reasonable mechanism based dynamics which would thus relieve them of making such a criticizable concession?
LikeLiked by 1 person
Eric,
Why do you think considering qualia as information processing is tricky? Can you identify qualia which are not information, or the processing of information? If you can, if you can provide examples that can’t be accounted for with information processing, then I’d have to concede that we need to identify additional mechanisms. Use your favorite example, thumb pain, and tell me what part of it transcends information processing.
LikeLiked by 1 person
Mike,
The reason for my “tricky” characterization of considering qualia to exist as information processing [and alone rather than mechanism instantiated, of course], is that from this premise there is no alternative except for the informationism position to be true. This is to say that it would inherently be possible for certain information on paper that’s converted into other information on paper, to create something that experiences what you know of as “thumb pain”. Thus some will understandably be self conscious about begging the question here — to begin this way is also to commit a fallacy.
On identifying qualia which exist without the benefit of information processing, yes, theoretically under the alternative to the informationist platform, this may occur. Unfortunately I’m unable to display an actual example for consideration, since the brain is all that I know of which produces qualia, and it’s a highly evolved information processing machine. But I am able to walk things back in a conceptual sense to provide a non-supernatural account of qualia as I see it. Let’s get into evolution.
At the beginning of the evolution of life, certain things must have replicated themselves (whether prions of whatever). The more successful of them could in a sense be considered “genetically superior”, since they would have had traits which helped them continue such replication. Regardless, entities with genetic material as we know it must have progressively emerged from this evolution. Because genetic material tends to accept input substances and process it algorithmically for output function, I consider this to be the world’s first “computer”.
Multicellular organisms would evolve, and thus each individual cell would be armed with its own computer / factory. Though the genetic material of each could be set up for general organism function, specialized “nervous” cells would evolve to connect input information to output organism function, inciting a given kind of movement or whatever.
Apparently these nerve cells began to be connected to a central place, permitting a wide range of input information to be taken together and algorithmically processed for more nuanced output function. Today this processing is considered to exist through the function of neurons and synapses. I thus consider this central processor as our world’s second variety computer.
Though fabulous non-conscious computers would then be free to evolve to provide more effective organism instruction, their essential weakness should have been the same weakness associated with the computers that we build. This is to say that they’d tend to fail under more “open” environments, and apparently because generic programming tends to not be specific enough to deal with these situations as effectively as they might. My proposal is that evolution was able to get around this by creating a strange teleological form of computer that rides on the backs of the other two.
If it isn’t information processing alone, then there must be mechanisms by which qualia become created. And for the sake of argument let’s theorize this as some form of electromagnetic radiation (even though we’d need experimental verification for that ultimately). Thus theoretically if EM waves were propagated in a laboratory with the exact right parameters, then something in it should thus experience what I do when my thumb gets whacked — the waves would exist as the qualia itself. Of course we’d use an information processing computer for such an experiment, though I see no reason to downplay an information processing element to qualia. Still it’s the mechanical instantiation that would render such an experiment natural, though as we’ve discussed I wouldn’t consider generic information alone to be as well.
So back to evolution, we left things such that organisms with incredibly advanced computers at their disposal, would still fail under more open environments given that it shouldn’t have been possible for them to be instantly reprogrammed to effectively deal with sufficiently novel situations. But at some point there should have also been EM fields based upon certain combinations of neuron firing that existed as qualia itself, or a functionless conscious entity. Given the serendipitous nature of evolution, it must have given those experiencers certain chances to make effective organism decisions based upon the good to bad that they would experience, and those choices must not have been poor enough for this form of function to die out. In fact these decisions must have been good enough for the experiencer to progressively be provided with more than base pain / pleasure types of experiences, like qualia in the form of memory, sight, hearing, and so on. This now functional third form of computer, is the medium by which you and I experience our existence.
Because psychology supervenes upon neuroscience, neuroscience should need psychology based accounts of our function in order for it to finally begin exploring consciousness effectively. Furthermore we don’t seem to yet have much success in psychology, given the failure of philosophy. My own ideas address each.
By the way, I wonder if you noticed Sabine’s new post stating that Einstein’s greatest achievement was the thought experiment? This was obviously a rhetorical hook, but clearly the thought experiment itself has been an invaluable tool for science. http://backreaction.blogspot.com/2020/07/einsteins-greatest-legacy-thought.html?utm_source=feedburner&utm_medium=email&utm_campaign=Feed:+blogspot/ermku+(Backreaction)&m=1
LikeLiked by 1 person
Eric,
“Thus some will understandably be self conscious about begging the question here — to begin this way is also to commit a fallacy.”
It’s not begging the question to follow logic to its conclusion. That would only be true if part of the justification assumed the conclusion. But you haven’t identified that. If you can identify it, or find some other flaw in the logic or its premises, then you should. Or show some evidence that trumps the logic. But simply refusing to accept the conclusion is illogical, an appeal to the stone, flinching from implication you dislike. (Unfortunately, not an uncommon situation in discussions of the mind.)
“This is to say that they’d tend to fail under more “open” environments, and apparently because generic programming tends to not be specific enough to deal with these situations as effectively as they might. ”
This assumes that standard computers are as good as it gets for intelligence. That seems questionable, and has the potential to be falsified as self driving cars and other autonomous robots increase in sophistication. I think they’re already nipping at its heels.
The rest, which you’ve gone over many times, is, I think, loose speculation, with multiple assumptions not driven by the data. It appears to be driven by a conviction that some form of physical dualism must be true.
Psychology supervening on neuroscience means psychological phenomena are higher level manifestations of neural mechanisms. If one’s going to cause the other to change, it will be the cause, not the effect.
I did see Hossenfelder’s posts. I’d call your attention to the last sentence.
We don’t hear about the many physics thought experiments that flounder at this step. And philosophical thought experiments lack it completely, which is why they’re only useful for clarifying intuitions.
LikeLiked by 1 person
Mike,
You asked “Why do you think considering qualia as information processing is tricky?”. I then displayed that from this premise it’s perfectly certain that the right information on paper that’s properly converted to another set of information on paper, must then create what I know of as “thumb pain”. Is it your conviction that the premise of “qualia as information processing” does not mandate that each and every instance of qualia could theoretically exist by means of some amount of information on paper that’s properly converted into another? I could understand someone making that argument, but to me the openness of paper as an information medium with arbitrarily fast processing mechanism, does seem to leave a mandated conclusion. Thus I see a begged question here. Just because you’ve been able to observe associations between information processing and qualia to build your statement, shouldn’t put it in the clear when expressed as strongly as you have.
Instead of beginning with qualia as information processing itself, you could say that all examples of qualia that we know of concern information processing. I do agree with that. And note that this leaves the potential for such processed information to also animate qualia producing mechanisms, whatever they might be, and so put the scenario in line with the function of all else that’s known in science.
Beyond question begging, I also argue that qualia as information processing alone would be supernatural. I know of nothing like this in nature. So I on my side I don’t see any “stone” present. I provide reason for my beliefs. I suppose that given our conflicting investments, we each see some “stone” on the other side.
On our computers, I of course agree with you that they’ll continue to improve. It’s just that evolution seems able to do things far better than us, and yet instead of continuing on with non-conscious organisms (perhaps ultimately creating something like an actual philosophical zombie rather than us), it added qualia as well. If F&M are right then this addition incited the Cambrian explosion around a half billion years ago. Self driving cars without qualia should do fine for our purposes, though for survival under more open environments I suspect that evolution needed the tiny third computer to instruct the massive second.
To me “physical” and “dualism” are contradictory terms. I consider the ideas of David Chalmers essentially “substance dualism”. Could you explain what it is about my ideas that you consider to (I guess) display a non natural theme?
The reason that I brought the supervenience of psychology into this, is because in order to understand how something works at a lower level, it makes sense to grasp it somewhat first at a higher level. Imagine science trying to reverse engineer a television, though without knowing anything about its purpose. Perhaps these scientists aren’t even furnished with the idea that the glass screen is for viewing. To me this would be kind of like trying to do the neuroscience on consciousness, though without the information which successful psychologists should some day provide regarding our nature.
On experiments, thought or otherwise, I’m certainly in favor of them. As I’ve mentioned several times, in order to agree upon a qualia instantiation mechanism we’d need experimental evidence. In the mean time, thought experiments can help clarify the implications of various proposals, as Sabine has mentioned. I’ve developed one which suggests that the informationism premise is supernatural for example. So that would be something for general consideration should my ideas ever gain any notable attention.
One thing about EM consciousness theory, is that it’s eminently testable. For example, let’s say that under the brain’s vast amount of neuron firing, some day a documented neural signature were found for someone regarding the qualia of a pin prick on the shoulder. Ah, so would this specific qualia exist because certain information is accepted and processed into other information? Or perhaps because the resulting neuron firing animates EM fields which constitute the person’s experience? Given the signature neuron firing of the pin prick, let’s say that we were to alter the associated EM waves with opposing waves and then see what the subject says about the experience. Does this now feel differently? If so then that would be strong evidence in favor of such radiation as a medium for consciousness. If nothing interesting results then we’d keep looking until something did finally get interesting.
LikeLiked by 1 person
Eric,
The thing about your stacked paper scenario is you tend to focus on just that, the stacked paper, and forget the processing involved, which I believe was a supercomputer in the latest version. The paper is just an information storage medium, no different from the transistors in computer chips, magnetized storage on hard drives, or the strength of synapses. Emphasizing just the stacked paper is misleading. We have to consider the whole system. (I would note having to store and retrieve information from the paper still involves issues similar to the Chinese Room, since even for a supercomputer, storing and retrieving it would be a major I/O bottleneck.)
On dualism, Chalmers’ dualism is property dualism, not substance dualism. https://en.wikipedia.org/wiki/Property_dualism
On why your idea is a type of dualism, you label it that way all the time: the first and second computer: dual computers. And the fact that the second computer is composed in some exotic manner (EM field interactions, or whatever) makes the second computer into a sort of photonic ghost. It’s like a sort of pseudo-substance dualism, but where the substances are neurons and EM fields.
And the assumption of dualism in your conception of qualia is so ingrained, you don’t even seem to realize it. As I’ve noted before, you see me not accounting for that dual nature, so you assume that my failure to do so means I’m sneaking in some supernatural version. The idea that there simply is no such component seems to be inconceivable for you.
LikeLiked by 1 person
Okay Mike, in the future for my thought experiment I’ll try to make sure to emphasize that the thumb pain would occur by means of the processing of the information, not just as stacks of information laden paper themselves. And furthermore that the supercomputer which does this processing might need to be insanely fast given that the paper medium could present serious bottlenecks. It’s a thought experiment and so merely requires one to think about this stuff properly.
On Chalmers, I still see his “property dualism”, or “naturalistic dualism”, as substance dualism. Given my own brand of metaphysics, that’s all still supernatural.
On the dualism of my “dual computers” model of brain function, I had no idea that you were talking about that. Yes I do make a distinction between the non-conscious function of most everything, including brains themselves, and the consciousness which can be produced by the brain. Thus dual computers, or even “dualism” I suppose. If characterized this way it would constitute a form of dualism that actually is natural.
As I see it the informationism approach could be characterized under my dual computers architecture as well. Let’s say that the brain does not happen to be processing any information for producing qualia at a given moment. Thus as I define it there’d be no conscious entity, but rather just a functional brain doing its standard non-conscious processing. Perhaps the subject is under the effects of anesthesia?
Regardless, when the brain does begin processing various forms of information associated with vision and so on, and so produces the qualia which I associate with consciousness, a “second computer” could be said to exist by means of this information processing. I consider this conscious entity to accept three forms of input information (qualia, non-qualia senses, and memory), to interpret them and construct associated scenarios (or think), resulting in decisions (some of which could be to operate various muscles). So my psychology based dual computers model of brain function does still work under the informationism platform. I just don’t consider it possible for information processing without instantiation mechanisms to produce anything, including qualia.
It’s not that I’m trying to take a line against your beliefs specifically, but rather an entire system which I consider quite messed up. Given that I’m a nobody, for my ideas to ever get anywhere I’ll need to come up with some kind of “in”. Viewed in retrospect, given the massive mistakes made by John Searle, it’s possible that I could make a name for myself by effectively demonstrating that the informationism platform depends upon a supernatural premise. Even if you always oppose me on this, I suspect that you’d also be happy for me if my project ever did get anywhere.
LikeLiked by 1 person
Eric,
Speeding up the computer doesn’t help. It would just spend more time waiting for the paper to be scanned or printed. You need to speed up the storage medium. Of course, the more we adjust along these lines, the less it flatters the anti-computationalist intuition.
It’s interesting that people object when I point out issues with these thought experiments, that I’m not thinking “about this stuff properly.” It only shows how much these are really just rhetorical devices.
There are plenty of people out there who oppose computationalism and will cheer any attempt to take it down. Searle’s views are actually quite popular in this camp. But if you want to convince any actual computationalists, your argument is going to have to be a lot more cogent. As things stand at the moment, I find it a pretty confused jumble.
I’ll be happy for any success you achieve, but balanced by sadness if it’s based on ideas I think are mistaken.
LikeLiked by 1 person
Mike,
I’d finally like to take your earlier suggestion for us to “agree to disagree”. Apparently each of us are too convinced, and on a matter that we find too important, for movement to be made by either of us. Hopefully we’ll be able to bracket this issue and go on to different interesting discussions. For this issue however I’d hope to find other intelligent people to have discussions with, whether also in the informationist camp, or undecided, or certainly people who are essentially on my side though aren’t yet familiar with my arguments.
I’ll always cheer your successes as well, and even if they lie in opposition with some of my most cherished ideas. Though it may be difficult for me personally to grasp, I suppose it’s still possible that my ideas aren’t actually all that good…
LikeLiked by 1 person
Thanks Eric. Looking forward to our other conversations. I’ve done a number of posts after this one you might find interesting.
LikeLiked by 1 person
Wer ist Kurz, und wen interessiert es, was er gesagt hat? (Who is Kurz, and who cares what he said?)
Sorry, couldn’t resist. Actually their discussion of intelligence struck me as reasonable, as far as it goes.
LikeLiked by 1 person
You made me look up what “Kurzgesagt” means: “In a nutshell.” (I have no idea why they did it in German.)
LikeLike
Literally: short said
LikeLiked by 1 person
I think of intelligence as consisting of 3 things:
1- varying actions/understandings or potential of varying actions/understandings to situations and stimuli
2- selection of best actions/understandings to serve the interests of the actor
3- persistence or memory of the selections so that new actions/understandings can be built on prior ones
This has all of the characteristics of unlimited associative learning but also the characteristics of Universal Darwinism, which may be why evolution appears to be a product of intelligent design.
LikeLiked by 1 person
I think those are good criteria. But I also think the ones in the video are too: gathering information, memory, and learning. My overall criteria is the ability to make accurate predictions, at least accurate enough for the system’s goals. In particular, learned predictions, since predictions made by instinct / programming aren’t really predictions by the system itself.
LikeLiked by 1 person
I’m not sure that ” gathering information, memory, and learning” are a lot different from mine. I just framed my list in manner similar to the requirements for universal Darwinism to make the point that intelligence does not require consciousness or even necessarily living organisms if the same characteristics can be implemented in a computer or some other sort of hardware. Criteria 2 is the problematic one for non-living organisms. For living organisms, it is fitness that is the selection mechanism. For conscious living organisms, it is also fitness in the end but there can be a lot of intermediate interests with no obvious fitness function and actions can be so much more nuanced. We might work to eat to survive – a fitness function; but that doesn’t stop us from quitting a boring, unsatisfying job for a more satisfying one with less pay which theoretically might result in less fitness. I suspect it is complexity of the nuances of criteria 2, how we judge our interests, that relate to general intelligence and creativity but I am not sure how those things get put into AI. I may prefer red and you may prefer blue but what would AI prefer and why would we make it one way or the other?
LikeLiked by 1 person
Definitely there’s a lot of overlap between your criteria and the video’s.
I actually misread Criteria 2 the first time, thinking you meant action selection on the part of the system itself, but you actually meant selection of actions the system is capable of. Certainly for living systems, evolution makes that determination. For a non-living system, the designer would have to. Although we might imagine an iterative process of numerous prototypes, where the available actions are adjusted with each new version, which seems like it might fit under he universal Darwinism concept.
The nuances question is interesting because we actually don’t always make the most adaptive choice. But that fits when you remember that our innate preferences evolved under various circumstances and for various reasons that might not exist anymore. We’re no longer living in bands of hunter gatherers on the African savanna. (And many of impulses go back even further to our tree swinging ancestors.) So often we’re going to make decisions for reasons that may not tie back to ultimate adaptability. We could view it, strictly from an evolutionary perspective, as impulses that were once adaptive misfiring in the current environment.
LikeLike