Neurologist outlines why machines can’t think

[devolved]

I base it on what can be observed and tested. Because I have no other choice.
If you think there is another valid basis to work from, I'm all ears.

Do you or don't you assume that what you see is material based on properties you ascribe to things? I understand that your assumptions and conclusions are driven by observations, but I'm not interested in your observation as much as I'm interested in your assumptions and conclusions.

We can be performing the same observation, but arriving at different conclusions and assumptions. That's what you continually ignore in this conversation when you insist that yours and only your assumptions about reality matter, especially when you assume that reality is "material" in nature and then you go on to circularly demand the evidence for immaterial. Of course if you assume that everything you looking at is material... there's no evidence for immaterial that can be presented to you.

The question is how do you know that what you are looking at is material and not a manifestation of "immaterial"?

 

[FrumiousBandersnatch]

... I'm speaking about this from personal "meta-cognition" perspective, and it seems like our conscious experience "waits" on those things before these pop into awareness.

If you get to the point of saying that everything that the meta-cognitive 'I' (the waiter-on-experience), experiences, including its sense of awareness and self, is consistent with internally generated phenomena, then I wonder what is there when those phenomena are absent; it seems to me that it doesn't exist without those internally generated phenomena, i.e. when there's nothing to be aware of. This suggests to me that the meta-cognitive 'I' is, in some sense, the integration of these experiential phenomena.

Kahneman's System I (unconscious) & System II (conscious) description, and the neurological observation that processing of stimuli below the threshold for conscious awareness remains local to the relevant sensory processing area, and that the processing of stimuli above that threshold extends beyond the local to involve widespread brain activity, and the apparent correspondence between richness or level of conscious experience and degree of integration of information (processing) described by Tononi et al, are consistent with the idea that the metacognitive 'I' is the integrated widescale processing of preconscious outputs (including outputs associated with the sense of self) across the brain.

Why it feels like something to be a system performing that kind of processing, I don't know - and I can't see how we can answer the question beyond discovering the differences between the configurations that do and don't have subjective experience; and that itself is a major challenge if some configurations may have subjective experience but cannot report it.

 

[Silmarien]

Sure; my point was that intuition can assist in evaluating the validity of the argument, particularly with lengthy or complex arguments. We don't have to accept the truth of the premises to do this.

Provisional 'for the sake of the argument' acceptance of premises is common, and it's not unheard of for an argument to have sufficient explanatory power to change one's estimate of the likelihood of an uncertain premise being true - it happens in science a fair bit.

You still haven't addressed my point at all. What is intuition, what is a priori knowledge, what is logic, and why should we care about what happens in science if we can't answer these questions in the first place? You're side-stepping the epistemological issues entirely.

 

[zippy2006]

Sure; my point was that intuition can assist in evaluating the validity of the argument, particularly with lengthy or complex arguments. We don't have to accept the truth of the premises to do this.

Provisional 'for the sake of the argument' acceptance of premises is common, and it's not unheard of for an argument to have sufficient explanatory power to change one's estimate of the likelihood of an uncertain premise being true - it happens in science a fair bit.

Intuition may be helpful in evaluating an argument, and provisional acceptance of premises can be useful, but some intuitions are more foundational and necessary than that. For example, some of the most foundational would be things like the principle of non-contradiction and the principle of sufficient reason.

It seems to me that Silmarien was saying that once you recognize the value of some intuitions you must be prepared to distinguish true intuitions from false intuitions rather than simply apply blanket skepticism to intuitions in general. The same applies to a priori knowledge.

 

[expos4ever]

Convinced of what precisely? That there is a problem with materialism or did you adopt Chalmer's panpsychism?

I find the argument that qualia cannot be reduced to structure and function - which is really what our standard models of reality are all about - to be intuitively overpowering. This is my response to your question about materialism - I am not sure exactly what you mean by that term.

As for panpsychism, I think Chalmers arguments are hard to refute, even though, like others, the whole idea seems a tad crazy.

 

[devolved]

If you get to the point of saying that everything that the meta-cognitive 'I' (the waiter-on-experience), experiences, including its sense of awareness and self, is consistent with internally generated phenomena, then I wonder what is there when those phenomena are absent; it seems to me that it doesn't exist without those internally generated phenomena, i.e. when there's nothing to be aware of. This suggests to me that the meta-cognitive 'I' is, in some sense, the integration of these experiential phenomena.

Consciousness is both an awareness of environment and awareness of being an observer. So absence of environment doesn't necessitate absence of observer in context of awareness of the awareness.

I don't know what that would be like in context of lack of sensory input. It's a good question. But I'm not sure that we can conclude that such awareness would not exist given the lack of "content".

I agree that feedback is a problem.

 

[FrumiousBandersnatch]

You still haven't addressed my point at all. What is intuition, what is a priori knowledge, what is logic, and why should we care about what happens in science if we can't answer these questions in the first place? You're side-stepping the epistemological issues entirely.

My post was an explanation of what I meant by my previous post, not an exploration of epistemological issues.

 

[FrumiousBandersnatch]

Intuition may be helpful in evaluating an argument, and provisional acceptance of premises can be useful, but some intuitions are more foundational and necessary than that. For example, some of the most foundational would be things like the principle of non-contradiction and the principle of sufficient reason.

It seems to me that Silmarien was saying that once you recognize the value of some intuitions you must be prepared to distinguish true intuitions from false intuitions rather than simply apply blanket skepticism to intuitions in general. The same applies to a priori knowledge.

That seems intuitively reasonable

 

[Silmarien]

My post was an explanation of what I meant by my previous post, not an exploration of epistemological issues.

If you want to call out other people for relying upon intuition incorrectly, you really ought to explore the epistemology involved. Otherwise, there's going to be a lot of special pleading involved in your own account of appropriate places to use intuition.

 

[FrumiousBandersnatch]

Consciousness is both an awareness of environment and awareness of being an observer. So absence of environment doesn't necessitate absence of observer in context of awareness of the awareness.

That's why I said, "If you get to the point of saying that everything that the meta-cognitive 'I' ... experiences, including its sense of awareness and self, is consistent with internally generated phenomena".

There's a good neurological case for the 'self' being a construct assembled from pre-conscious processes that map into the conscious arena 'images' of:

a perspective or viewpoint in which non-self objects are mapped (i.e. with respect to the body);
a sense of ownership of the mind doing the mapping;
a sense of agency (ownership of bodily actions);
'primordial feelings' (originating in interoception) that are independent of engagement with external objects.
​

The combination of the aggregate self and the non-self object images makes a conscious mind - suggests Antonio Damasio in 'Self Comes to Mind'. Here consciousness is a particular state of mind with a 'self' process added to it. That state of mind is a dynamic set of images of non-self objects mapped in the mind together with the self image, assembled as above.

I don't know what that would be like in context of lack of sensory input. It's a good question. But I'm not sure that we can conclude that such awareness would not exist given the lack of "content".

I wonder what awareness is, if not awareness of something?

 

[FrumiousBandersnatch]

If you want to call out other people for relying upon intuition incorrectly, you really ought to explore the epistemology involved. Otherwise, there's going to be a lot of special pleading involved in your own account of appropriate places to use intuition.

I don't want to call out other people for relying upon intuition incorrectly. I suggested a context where I think it is likely to be most reliable. YMMV.

 

[Ana the Ist]

This article is an interesting discussion about why it is not AI that is a threat to humanity. Rather it is humanity that may use AI to destroy itself.
https://mindmatters.today/2018/neurosurgeon-outlines-why-machines-cant-think/
A cornerstone of the development of artificial intelligence is the pervasive assumption that machines can, or will, think. Watson, a question-answering computer, beats the best Jeopardy players, and anyone who plays chess has had the humiliation of being beaten by a chess engine. (I lose to even the most elementary levels of the chess program on my iPhone). Does this mean that computers can think as well as (or better than) humans think? No, it does not. Computers are not “smart” in any way. Machines are utterly incapable of thought.

The assertion that computation is thought, hence thought is computation, is called computer functionalism. It is the theory that the human mind is to the brain as software is to hardware. The mind is what the brain does; the brain “runs” the mind, as a computer runs a program. However, careful examination of natural intelligence (the human mind) and artificial intelligence (computation) shows that this is a profound misunderstanding.

What is the hallmark of human thought, and what distinguishes thoughts from material things? Franz Brentano (1838–1917), a German philosopher in the 19th century, answered this question decisively. All thoughts are about something, whereas no material object is inherently “about” anything. This property of aboutness is called intentionality, and intentionality is the hallmark of the mind. Every thought that I have shares the property of aboutness—I think about my vacation, or about politics, or about my family. But no material object is, in itself, “about” anything. A mountain or a rock or a pen lacks aboutness—they are just objects. Only a mind has intentionality, and intentionality is the hallmark of the mind.....

....But to believe that machines can think or that human thought is a kind of computation is a profound error. Belief in this fundamental error about AI will lead us away from, not toward, the truth about AI. Machines, for example, will never become malevolent and harm mankind. Men will act with malevolence, using machines, or men will use machines in ways that (unintentionally) harm others. Men can use cars malevolently and carelessly and can thus harm others. But the malevolence and careless is in the man, not in the car.

To paraphrase Pogo: we have met AI, and AI is us.

By Denyse O'Leary

Interesting theory.

 

[zippy2006]

My concern is actually that sufficiently advanced artificial intelligence might possess self-consciousness and subjectivity. There are so many different metaphysical approaches to the mind, and even if we want to rule out materialism, there are alternatives out there that also support the possibility of genuinely conscious AI.

I'm specifically thinking about panpsychism and idealism, both of which approach subjectivity as something that is ontologically fundamental and/or ubiquitous. Arch anti-materialist David Chalmers is a serious proponent of AI, so this is not the sort of thing that sets people in easily predictable groups.

I could really go either way, since I'm undecided on what Intellect actually is in terms of the human mind. (I'm somewhat unconvinced by the entire Western tradition on this issue, from Plato and Aristotle to the modern materialists.) If we were able to create an artificial brain whose neural network perfectly mimicked the human brain, would it be conscious in a similar sense? We could drop materialism for a form of idealism wherein the physical serves to facilitate the manifestation of the underlying subjective reality, and there would still be no real reason to assume that this couldn't be done artificially.

Like I said, it's Vedic thought that I find more seductive than I would like on this particular issue, and once you toss that into the mix, things become considerably more difficult.

Right, and is your Vedic conception of Intellect compatible with self-conscious AI on the supposition that Intellect is necessary for self-consciousness? I am not too familiar with Hinduism. What precisely about the Western tradition of Intellect do you find lacking?

I will have to read more about panpsychism. Is it more than a mere solution to 'the interaction problem'?

(Interestingly, I tend to find reductive idealism and reductive materialism to be two sides to the same coin, especially on questions like this one.)

I've noticed that as well. Presumably you also find them both lacking. What do you think it means? Does it present a case for some kind of dualism?

Yes, I have similar concerns. There are some very good materialists out there, and I could certainly make a theistic physicalist theory of mind work, but by and large it just seems to be a cultural prejudice. Though it was actually New Mysterianism that made me stop wondering if I was missing something as far naturalism was concerned--if you're going to simultaneously hold that we are evolutionarily unequipped to tackle the question of consciousness and that it must have a naturalistic explanation, we are clearly into the realm of dogma.

I agree, but are naturalists in the habit of claiming that we are evolutionarily unequipped? Maybe they should be, but that's beside my question.

Oh, it gets worse. According to the forum, I am a radical Marxist with a conservative religious worldview.

You are going to be a great bridge-builder someday.

You know, I have Caputo's book on Kierkegaard, but I haven't read it yet. I'd heard about his book on Heidegger also, but my reading list grows a lot more quickly than I can clear it. I do have some stuff from within the Christian world in a similar trend--David Bentley Hart's Beauty of the Infinite is suppose to hold up Maximus the Confessor as a response to Heidegger, though it'll be a while before I get around to it since I'm meaning to read Hans Urs von Balthasar's Cosmic Liturgy: The Universe According to Maximus the Confessor first. (Just don't tell the Orthodox, haha.)

Apparently Caputo has a number of books on Heidegger. I am aware of Hart's book and it is something I would like to read at some point. You should drop me a line when you finish either of those to give a brief assessment.

See, in a way you're dusting off my dormant neurons. In the last few years I've been more interested in mysticism and psychology than, say, philosophy and theology (even though there are clear overlaps). That's actually what brought me back to CF, but I only found one related thread. Yet I really admire your intellectual verve and it's fun to revisit such things. Further, my question about Neoplatonism was inevitably related to my current interests, as it tends to implicate mysticism more than Aristotelianism and Scholasticism do.

But yeah, Heidegger is... very difficult. I actually need to reread him--I can't even say whether or not the university course where I'd studied him was subpar, since I managed to forget which one it was altogether. It took finding the essays to remember where I'd come across him at all!

lol!

Hahaha, you've left me wondering which one!

When I signed up they gave me this blanket of anonymity and I just find it to be so cozy!

 

[Justatruthseeker]

Actually dominance hierarchies is how any society operates and organizes. You can't have organization without dominance hierarchies. Lack of dominance hierarchy would actually be closer to anarchy than a strict dominant hierarchy.

Dominance isn't inherently evil or unwanted. Dominance hierarchies is how we operate in this world, and there's nothing wrong with these apart from immoral settings of dominance, which is much more rare than that which is moral form of dominance.

Hence, we really have to focus on contextual morality and not dominance. Dominance is merely a continuum of one's action to successfully alter reality, and it can exist on a short and wider spectrum of "influence over environment and people that occupy it". Thus, dominance is a generic concept. Morality is generally a very specific concept.

For example, it's wrong to kidnap people. Such would be a very specific context of the wrong type of dominance that we can contextualize and define.

Hitler’s Germany was a dominant society. The Church's inquisition was a dominant society.

Democracy is less dominant in that all have a say to a degree.

Kidnapping is a human thing not practiced in the rest of nature. So can not be used as an example against nature....

 

[Justatruthseeker]

This article is an interesting discussion about why it is not AI that is a threat to humanity. Rather it is humanity that may use AI to destroy itself.
https://mindmatters.today/2018/neurosurgeon-outlines-why-machines-cant-think/
A cornerstone of the development of artificial intelligence is the pervasive assumption that machines can, or will, think. Watson, a question-answering computer, beats the best Jeopardy players, and anyone who plays chess has had the humiliation of being beaten by a chess engine. (I lose to even the most elementary levels of the chess program on my iPhone). Does this mean that computers can think as well as (or better than) humans think? No, it does not. Computers are not “smart” in any way. Machines are utterly incapable of thought.

The assertion that computation is thought, hence thought is computation, is called computer functionalism. It is the theory that the human mind is to the brain as software is to hardware. The mind is what the brain does; the brain “runs” the mind, as a computer runs a program. However, careful examination of natural intelligence (the human mind) and artificial intelligence (computation) shows that this is a profound misunderstanding.

What is the hallmark of human thought, and what distinguishes thoughts from material things? Franz Brentano (1838–1917), a German philosopher in the 19th century, answered this question decisively. All thoughts are about something, whereas no material object is inherently “about” anything. This property of aboutness is called intentionality, and intentionality is the hallmark of the mind. Every thought that I have shares the property of aboutness—I think about my vacation, or about politics, or about my family. But no material object is, in itself, “about” anything. A mountain or a rock or a pen lacks aboutness—they are just objects. Only a mind has intentionality, and intentionality is the hallmark of the mind.....

....But to believe that machines can think or that human thought is a kind of computation is a profound error. Belief in this fundamental error about AI will lead us away from, not toward, the truth about AI. Machines, for example, will never become malevolent and harm mankind. Men will act with malevolence, using machines, or men will use machines in ways that (unintentionally) harm others. Men can use cars malevolently and carelessly and can thus harm others. But the malevolence and careless is in the man, not in the car.

To paraphrase Pogo: we have met AI, and AI is us.

By Denyse O'Leary

Oh agreed. AI could make our lives a paradise. But there will always be another Hitler that would warp the AI to its own goal to dominate and subjugate all others...... As long as man could ultimately control the machine, the machine could be used to his own ends.....

 

[devolved]

Hitler’s Germany was a dominant society. The Church's inquisition was a dominant society.

You are conflating dominance with oppression. The above were both dominant and oppressive.

You parents exercised dominance over you. Any authority would be dominant. Governments are by their nature a dominant force.

Democracy is less dominant in that all have a say to a degree.

First of all you are only talking about political dominance, which democracy as a system doesn't prevent. Yes, individuals have a say, but that say is in a form of a representative vote, in which case it becomes a form of pluralistic dominance.

Democracy is a statistical survey and reinforcement of dominant opinions.

Either way, that merely one of the dominant places on the dominance hierarcy that you are looking at. Economic dominance can be much more influential than political one, since it can be more permanent.

Kidnapping is a human thing not practiced in the rest of nature. So can not be used as an example against nature....

It's not. Sea otters, for example, kidnap and hold the kids as a random for food.

I'm not really sure why you say the above to begin with, since your original point was that there is more dominance and cruelty in "nature".

Are humans a part of nature?

 

[Justatruthseeker]

You are conflating dominance with oppression. The above were both dominant and oppressive.

You parents exercised dominance over you. Any authority would be dominant. Governments are by their nature a dominant force.



First of all you are only talking about political dominance, which democracy as a system doesn't prevent. Yes, individuals have a say, but that say is in a form of a representative vote, in which case it becomes a form of pluralistic dominance.

Democracy is a statistical survey and reinforcement of dominant opinions.

Either way, that merely one of the dominant places on the dominance hierarcy that you are looking at. Economic dominance can be much more influential than political one, since it can be more permanent.



It's not. Sea otters, for example, kidnap and hold the kids as a random for food.

I'm not really sure why you say the above to begin with, since your original point was that there is more dominance and cruelty in "nature".

Are humans a part of nature?

Then kidnapping would be a part of nature and mans morality against it in opposition to nature.

Are humans a part of nature? Yes. Although we tend to change nature to suit us, not change to fit nature.

If man is part of nature then dogs would be the longest running test of evolution, dispelling the notion that one species changes into another and simply confirming variation “within” the species. People can’t have it both ways......

 

[devolved]

That's why I said, "If you get to the point of saying that everything that the meta-cognitive 'I' ... experiences, including its sense of awareness and self, is consistent with internally generated phenomena.

There's a good neurological case for the 'self' being a construct assembled from pre-conscious processes that map into the conscious arena 'images' of:

a perspective or viewpoint in which non-self objects are mapped (i.e. with respect to the body);
a sense of ownership of the mind doing the mapping;
a sense of agency (ownership of bodily actions);
'primordial feelings' (originating in interoception) that are independent of engagement with external objects.
The combination of the aggregate self and the non-self object images makes a conscious mind - suggests Antonio Damasio in 'Self Comes to Mind'. Here consciousness is a particular state of mind with a 'self' process added to it. That state of mind is a dynamic set of images of non-self objects mapped in the mind together with the self image, assembled as above.

There's also good psychological / neurophysiological case for self being a different type of "socially-reflexive process". So, just like the pain reflexes exist as immediate behavior, the construction of "social self" is a mechanism that takes and groups collection of reflexive behaviors that are invoked in a scope of certain cognitive context.

There's a good string of research to back that up from Russian neurosceinece that focused on behavior predominately from "reflexive behavior" standpoint.... beginning with Pavlov, and ending with Savelyev today.

So, this is a concept of "physical brain process of self" that I can readily concede, given what I see in research. But, that's not what I'm talking about when I'm talking about "conscious self". The "reflexive self" seems to be just that - a constructed set of behavioral preferences as a reflexive behavior in certain social or environmental contexts.

BUT, the above is not what I'm talking about when I'm talking about a process or "state of being" that exists and can be made aware of the mechanism of "self". You can actually induce this dissociation state by some form of prolonged sensory deprivation... like meditation is a dark and silent environments. When you exit that state, you are usually hyper-aware of various phenomenons of your perception that you don't generally focus on. I tend to notice plentiful things that I generally ignore.

That process can also numb the reflexive mechanism of "social self" so you are more aware of certain reflexive social absurdities we perform day to day.

I wonder what awareness is, if not awareness of something?

Well, awareness could be an inalienable property of or for existence. It could be just "is". You wouldn't say "I wonder what existence is, except for existence of something". Of course! .

If what we call matter reacts with other matter and doesn't with other, could we say that it has some embedded primordial intelligence as a part of that reactive or non-reactive mechanism?

 

[FrumiousBandersnatch]

There's also good psychological / neurophysiological case for self being a different type of "socially-reflexive process". So, just like the pain reflexes exist as immediate behavior, the construction of "social self" is a mechanism that takes and groups collection of reflexive behaviors that are invoked in a scope of certain cognitive context.

There's a good string of research to back that up from Russian neurosceinece that focused on behavior predominately from "reflexive behavior" standpoint.... beginning with Pavlov, and ending with Savelyev today.

If you have any references, I'd be interested; this would appear to be another case of forwarding the output of unconscious or preconscious activity (the reflexive behaviours) into the conscious arena (the 'Global Workspace'), triggering the widescale activation that signals awareness, with its accompanying sense of ownership and agency.

BUT, the above is not what I'm talking about when I'm talking about a process or "state of being" that exists and can be made aware of the mechanism of "self". You can actually induce this dissociation state by some form of prolonged sensory deprivation... like meditation is a dark and silent environments. When you exit that state, you are usually hyper-aware of various phenomenons of your perception that you don't generally focus on. I tend to notice plentiful things that I generally ignore.

That process can also numb the reflexive mechanism of "social self" so you are more aware of certain reflexive social absurdities we perform day to day.

Some people have said they get similar effects to what you describe from meditative practices, which affect the activity of the Default Mode Network and train the focus of attention.

From my own experience, it sounds like a similar experience (though, I should think, much subtler) to that from psychotropic drugs (e.g. psilocybin) that change the balance of neural activation & suppression and allow more cross-signalling between adjacent areas than is usual. This can cause the sensation of seeing things differently, or for the first time, or the focusing of attention on details normally filtered out and ignored, and various other changes and distortions in perception, cognition, and experience. The effects vary with the drug according to the neurotransmitters most affected and the feedback (via Kahneman's 'primordial feelings') from other physiological effects such as changes in hormonal secretion, etc.

I don't see anything in what you describe that could not easily be attributed to such changes in brain activity. I get the feeling you're saying that it feels as if there's more to it, a separate 'core of awareness' that gets to experience these things, but it seems to me that's what all these preconscious process outputs produce when integrated across the brain, all the resulting activity produces the sense of awareness.

I'm wary of ending up in a Dennetian 'Cartesian Theatre', where the homunculus of awareness sits in the head, perceiving all the information presented, but with nothing to explain how (never mind the infinite regress that results if one tries to explain it by proposing a homunculus of awareness inside the first one).

Well, awareness could be an inalienable property of or for existence. It could be just "is". You wouldn't say "I wonder what existence is, except for existence of something". Of course! .

An easy answer to the hard question?

If what we call matter reacts with other matter and doesn't with other, could we say that it has some embedded primordial intelligence as a part of that reactive or non-reactive mechanism?

Not quite sure what you mean, but we do know what we're made up of - protons, neutrons, and electrons, and the forces that are relevant at human scales - gravity & electromagnetism; and we know very precisely how they interact. There may well be other unknown particles and forces, but equally, it's clear that they must be too weak or too short range to be significant - if you accept our best model of how the world works - one that underlies our technology, is supported by literally billions of experiments, and has made valid predictions to the limits of our observational capabilities - see Sean Carroll's video, starting at 33 mins:

I might consider the idea of 'awareness' being somehow a fundamental attribute or property if someone came up with a precise characterisation of it, i.e. precisely what is meant by it, and a plausible hypothesis to explain how it fits with our other observations better than the worst hypothesis, 'it's just magic'

 

[Jesus_is_Saint]

it is like asking can a gun itself fire at us?