Conscious Entities

I see that while I was away the Internet has been getting a certain amount of stick over the way it allegedly alters our mental processes for the worse. Some of this dialogue apparently stems from a two-year-old piece by Nicholas Carr, now developed into a book.  Most of the criticisms seem to be from people who have experienced two main problems: they’re finding that they have a reduced attention span, and they’re also suffering from a failing memory.  They attribute these problems to Internet use – but I wonder whether they have made sufficient allowance for the fact that both can also be the result of simple ageing.

I think it’s true that if you don’t use your memory, it gets worse, so it’s superficially plausible that relying on the Internet could have a bad effect: but I don’t think I find myself using the Internet for things I would otherwise have learnt by heart, while I certainly have begun forgetting things I knew quite well before the Internet was invented.  So far as my attention span is concerned, it has certainly waned steadily over the whole course of my life: when I was four or five I could spend a long time just examining the patterns made by the grain in a piece of wood (mind you, in those days, we had interesting wood, not like the bland stuff they produce these days…).  I regret this to some extent, but in another way I don’t regret it at all, because I think it is partly a result of mental improvement, the result of accumulated experience. I can tell more quickly now when something is not going to be worth pursuing, and I am less bothered about dropping it quickly. Nowadays, I don’t feel at all guilty about dropping a book after chapter one if reading it looks like a mistake, whereas twenty years ago I would no more have stopped reading a book once started than I would have got up and left a dinner party half-way through. I know now that life is too short.

But there are deeper criticisms of the malign effects of the Internet.  Jaron Lanier, in an NYT piece (he too has written a book about it; it’s interesting that both he and Carr, in spite of the alleged waning of attention spans, still thought this quixotic ‘book’ business was still worthwhile, rather than just tweeting their thoughts), suggests that we are increasingly deferring to computers, encouraged by inflated claims made for various pieces of software. This has a serious moral dimension – if we see computers as people, we may be led to see people as mere machines – but it also undermines original creativity, a point developed more in the book (OK, I skimmed a few summaries). We start to value mashups and compilations as more valuable than new work generated from scratch. Perhaps worst of all, we may end up letting stupid algorithms make actual decisions for us.

The first of these points is one that has been made before, and I believe it underlies many people’s aversion to the whole idea of AI.  I think it’s undeniable that software producers are gravely inclined to overstate what their programs do, speaking of relatively simple data manipulation as though it involved genuine understanding and originality. But I don’t think that has really devalued our conception of humanity – not yet, anyway. Unless and until someone produces a machine which they claim is a conscious being, that remains a danger rather than a current problem. I don’t think we’re really in danger of delegating important decisions either; letting a computer suggest a track or a book is akin to random browsing of shelves; Lanier himself notes that even the advocates of the computers don’t allow the machines to design their products or run their companies.

There’s certainly something in the point about creativity. Hypertext encourages quotation, and I suspect that this has had an influence:  an apposite quote is a frequent and respected way of contributing to discussions on popular forums and blogs, to an extent that would seem almost donnish if the quotes weren’t typically from Star Wars or the Simpsons rather than Shakespeare. It must surely be the case that sometimes on the Web people use text quotes, photoshopped images and so on when otherwise they would have chosen their own words or drawn their own pictures; but mostly the copied stuff is surely extra. It’s a bit like photography; when people could take photographs, they made fewer engravings and oil paintings, but mainly they made many more pictures (and let’s be honest – some of those uncreated paintings and engravings were no loss).

There are deeper issues still: has the Web influenced the way we perceive the world? I strongly suspect that films have to some degree influenced the way I see the world and represent my own life to myself. I can’t be the only person who has sometimes felt an irresistible urge to do a reaction shot for the benefit of a non-existent audience (one day it may exist if CCTV continues to spread). In one way I think the Internet may have a more pervasive effect.  I remarked that the Internet is quotation-driven: but it doesn’t just quote, it comments. You could say, I think, that the essence of Web culture is to display something (text, picture, video) and provide comment in parallel (I suppose I’m exemplifying this as I describe it). I suspect that as time goes on reality will come to seem to us like the thing presented and our thoughts like the comments.  Our consciousness may end up seeming like a set of lengthy footnotes. Perhaps David Foster Wallace was way ahead of us.

There have been a number of attempts recently to externalise consciousness, or at least extend it beyond the skull. In Out of Our Heads, Alva Noë launches a very broad-based attack on the idea that it’s all about the brain, drawing in a wide range of interesting research – mostly relatively well-known stuff, but expounded in a style that is clear and very readable. Unfortunately I don’t find the arguments at all convincing; I’m not unsympathetic to extended-mind ideas, but Noë’s clear and thorough treatment tended if anything to remind me of reasons why the assumption that consciousness happens in the brain looks so attractive.

I’m happy to go along with Noë on some points: in his first chapter he  launches a bit of a side-swipe at scanning technology, fMRI and PET, pugnaciously asking whether it is ‘the new phrenology?’ and deriding its limitations: this seems a useful corrective to me. But in chapter two we are brought up short by the assertion that bacteria are agents.  They have interests, and pursue them, says Noë; they’re not just bags of chemicals responding to the presence of sugar.  Within their limits we ought to accord them some sort of agency.

To me, proper agency requires an awareness of what acts one is performing and the idea that bacteria could have it at any level seems absurd. How did we get here? Noë’s case seems to be that the problem of other minds is effectively insoluble on rational empirical grounds; we can never have really solid reasons for believing anyone else, or any other entity, is conscious; yet we find ourselves unable to entertain seriously the idea that our fellow-humans might be zombies. This, he thinks, is because we have a kind of built-in engagement, almost a kind of moral commitment. He wants to extend this to life fairly widely, and of course if brainless bacteria can have agency, it tends to show that the brain is a bit over-rated.  I think he’s unnecessarily pessimistic about the evidence for other conscious minds; as a matter of fact books like his are pretty spectacular evidence; how and why would human beings produce such volumes, examining the inner workings of consciousness in minute detail, if they didn’t have it? But bacteria have yet to produce such evidence in their own favour.

Noë rests a fair amount of weight on experiments which show the remarkable plasticity of the brain: notably he quotes experiments on ferrets by Mriganka Sur. New-born ferrets had their brains rewired in such a way that the eyes fed into auditory, rather than visual cortex; yet they grew up able to use their eyes perfectly well.  This shows, Noë suggests, that no particular part of the brain is required for vision. That might be so, but that in itself does not show that no brain at all will do the job, and obviously it won’t: if the ferrets’ optical nerves had been linked with their teeth, or left dangling unattached, they would surely have been unambiguously blind. The belief that consciousness is sustained by the brain does not commit us to the view that only one specific set of neurons can do the job.  Noë explains, quoting an experiment with a rubber hand, how our sense of our selves and where we are can be moved around in a remarkably vivid way. For him, this shows that where the brain is doesn’t matter; but for others it seems equally likely to suggest that what the brain thinks is crucial and where our real hand is doesn’t matter at all.

Noë wants to claim that the frequently quoted thought-experiment of a brain in a vat, extracted from the body but still living and thinking, is impossible in principle (we know it’s impossible in practice, at least currently). He suggests that even if we did manage to sustain a brain artificially, the supporting vat, providing it with oxygenated blood and all the other complex kinds of support it would need, would actually amount to a new body. This nifty bit of redefinition is meant to show that the idea of a brain without a body will not fly. But the real point here is surely missed.  OK, let’s accept that the vat is a new body: that still means we can swap bodies while maintaining an individual consciousness. But if we keep the body and swap brains… it just seems impossible to believe that the consciousness wouldn’t go with the brain.

This perception seems to me to be the unshiftable bedrock of the discussion. Noë expounds effectively the case for regarding tools and even parts of the environment as parts of our mental apparatus; and he brings in Putnam’s argument that ‘meaning isn’t in the head’. But these arguments only serve to expand our conception of the brain-based mind, not undermine it.

My sympathy for Noë’s case returned to some degree when he discussed language. He notes that for Chomsky and others language seemed a miraculous accomplishment because they misconstrued it as an exercise in the formal decoding of a vast array of symbols. In fact, language is an activity rather than a purely intellectual exercise, and develops in a context, pragmatically.  I’d go along with that to a great extent, but while Noë sees it as proving that our decoding brain isn’t the crux of the matter after all, it seems to me it proves that decoding isn’t really what our brains are doing when they process (a word Noë would object to) language.

I sympathised even more with Noë when he attacked the idea that reality is, essentially, an illusion. If this were the case, the brain would be the all-powerful arbiter of reality (although it might seem that if the world is an illusion, the brain must be one too, and we should be dealing with a mind whose actual nature need not be pinkish biological glop). But he seemed to be back on weak ground when he concluded by taking on dreams. Dreams, after all, seem like the perfect evidence that the brain can produce conscious experience without calling on the senses or the body.  Noë argues that dreams are more limited than we think, that not all waking experiences can be reproduced in dreams, which are always shifting and inconstant. This might be true, but so what? If the brain can produce conscious experiences on its own – any conscious experiences – that seems to show that, with all caveats duly entered, the brain is still where it really happens.

It’s a well-written book, and for someone new to consciousness it would provide many excellent short sketches of thought-provoking experiments and arguments. But I’m staying in my head.

There’s an illuminating new piece in the  SEP about 17th century theories of consciousness. (via) Your first reaction might be ‘what 17th century theories of consciousness?’; the discussion in those days was framed rather differently and it typically requires a degree of interpretation to work out what philosophers of the period actually thought about ‘consciousness’.  In fact, according to Larry M. Jorgensen, who wrote the entry, the 17th century saw the first emergence of the concept of consciousness as distinct from conscience: in many languages the same word is still used for both.

Hobbes apparently sets this out quite explicitly (somehow this interesting bit must have passed me by when I read Leviathan because it left no impression on my memory); he has conscience originally referring to something which two people knew about (‘knew together’), and then metaphorically for the knowledge of one’s own secret facts and secret thoughts. Jorgensen tells us that the Cambridge Platonists had a role in developing the modern usage in English where ‘conscience’ refers to knowledge of one’s own moral nature while ‘consciousness’ means simply knowledge of one’s own mental content.

That idea, of having knowledge of one’s own mental content, seems to have a reflexive element – we know about what we know; and this was an issue for philosophers of the period, notably Descartes. For Descartes it was essential that my having a thought involved me knowing that I had a thought; but for some this seemed to suggest a second-order theory in which a thought becomes conscious only when accompanied by another thought about the first.  Descartes could not accept this: for one thing if knowledge of my own thoughts is not direct, the cogito, Descartes’ most famous argument is threatened. The cogito claims that I cannot possibly be wrong about the fact that I am thinking, but if the knowledge of my thought is separate from the thought itself this no longer seems unassailably true.

It seems that while Descartes accepted that awareness of our own thought required some sort of reflection, he denied that the reflection was separate from the thought. He said that [T]he initial thought by means of which we become aware of something does not differ from the second thought by means of which we become aware that we were aware of it.

This can’t help but seem a little like cheating - sneaking in an extra thought for nothing.  I think the best way to imagine it might be through analogy with a searchlight. We can swing the light around, illuminating here a building, there a tree, just as we can direct our conscious awareness towards different objects. Then Descartes might ask: do we need a second light in order to see the first light? No, of course not, because the light is already illuminated; if the light lights up other objects it must itself be illuminated (if perhaps in not quite the same way).

A surprising amount of Jorgensen’s exposition seems to be relevant to current discussions, and not solely because he is, necessarily, reinterpreting it in terms of modern concerns. In some ways I’m afraid we haven’t moved on all that much.

I suppose the zombies couldn’t have the film industry all to themselves for ever, and here it is: Qualia, the movie (via). I wondered at first whether this was something to do with Sony and their Qualia man Ken Mogi, but in fact it seems it is a small independent venture. I said ‘here it is’, but actually all we have for the moment is a trailer: it seems that the funds required (amounts which I expect wouldn’t cover one day’s catering budget on a Hollywood blockbuster) have been difficult to get together.

How do you make a film about qualia? (Ken Mogi would probably ask how you could make one without them.) I can’t quite decide whether getting ineffable qualities into a film is an amusingly quixotic endeavour or an admirable ambition. It seems all too likely that you would end up with either the talkiest, chin-strokingest film ever made; or an exciting dramatisation of the life of Mary the Colour Scientist. (Susan Blackmore suggested that students should act out this famous gedankenexperiment, after all, though how that would help still rather eludes me.) Actually there’s no reason why a film can’t at least raise genuine philosophical issues. I’ll always remember the Captain’s advice on how to deal with the malfunctioning bomb in Dark Star (“Teach it phenomenology”), and The Matrix is often credited with asking interesting questions – though sadly the red and blue pills were soon put aside so that the film could become a kung fu movie performed by people dressed as a Eurythmics tribute band (The Revenge Tourists?).

I haven’t found much information about the actual plot of Qualia, but it seems it has something to do with research which triggers or examines ghostly occurrences and disturbs someone’s complacent monist materialism. Nothing wrong with disturbing our dogmatic slumbers, of course. I like to think that at some stage a grave scientist will say “Sir! We’re detecting… phenomena.”

But the association with ghosts is not particularly welcome. I wonder whether this is another sign, like the use of qualia to buttress the theist case, that the hard problem potentially appeals to those who would like the world to be less scientific and more magical.  I hope not: I’d hate to see New Age shops selling qualia-enhancing crystals. Perhaps that’s just snobbery?  After all It’s legitimate to claim qualia as evidence for some kind of dualism, and some kind of dualism is what you might well be looking for if you wanted to provide ghosts with some respectable ontological underpinnings. Still, I can only look forward to the film with qualified enthusiasm .

Mostyn W Jones might seem to be leading with his chin a little when he offers us, in the JCS, “a clear, simple mind-body solution”. His “clear physicalism”  is meant to banish many of the “obscurities” (a favourite word) involved in other accounts, especially reductionist and functionalist ones.

One problem with functionalism, according to Jones, is that it requires us to believe that conscious sensations are multiply realisable. So long as it embodies the right functions, any physical thing can have consciousness. But functions are abstract; what have they got to do either with my physical brain or with my vivid actual experience? These relations are surely “obscure”. Another problem is that none of these reductive theories can deal with qualia (the inherent phenomenal redness of red, hotness of heat, etc). Once you’ve reduced consciousness to computation, to non-computational functions, or to anything similar, you can no longer explain why it is that real inward experience occurs, or what causal relations that experience has with anything, or how it contrives to have them. Jones sympathises with Strawson and Stoljar: he would like to be able to say that qualia are just the reality of experience: science gives an outside account, and qualia are how it looks from the inside.

In some ways this is an appealing position to take, but there are a number of pitfalls. If all you’re saying is that science is the third-person perspective and qualia are the first-person perspective, you’re just re-stating the problem: why is there a first-person perspective, anyway? And if qualia are just the reality, you have two options. One is to find an explanation of why they don’t crop up all over the place, but seem only to arise in the presence of appropriate brain functions; the other is to bite the bullet and say that in fact they do crop up all over the place, and that panexperientialism is correct. Strawson, if I’ve understood correctly, leans in this latter direction; Jones, in spite of his hostility to functionalism, seems to lean back the other way: he describes consciousness as a neural substance arising from highly active, highly connected neural circuits; so there still seems to be a broadly functional explanation of why these phenomena are confined to brains.

“Substance” is a treacherous word: in ordinary parlance it suggests a lump of stuff, simple physical matter: but in philosophy, especially older philosophy, it has a much more slippery meaning as a basic element, one of the things that remains when analysis is complete. By pursuing this idea of a substance as something unanalysable to its logical conclusion, Leibniz produced the bizarre relativistic ontology of the Monadology:  Jones certainly doesn’t mean monads, but what does he mean? He says “Clear physicalism avoids this obscurity by treating qualia as electrochemical substances that underlie observable brain activity and do work in brains”.  The combination of both underlying observable physics yet doing work – causal work, we assume – seems very problematic. It would be odd but compatible with physics if Jones were merely saying that qualia are aspects of the physical world that run along with – perhaps over-determine – the causal effects specified by physics: but that doesn’t seem to be quite it.

If we take a step back, the idea of making consciousness a physical substance of any kind seems difficult to accept; consciousness, after all, comes and goes, but matter is conserved. If we take a more flexible view of the notion of substance and allow the substance of consciousness to come and go in line with certain vigorous firings of interlinked neurons it becomes a little hard to see what our quarrel with some sort of functionalism can be. Pain, it turns out, is not the firing of C-fibres (or whatever), but it is a substance that occurs in perfect correlation with the firing of C-fibres. Hmm.

Paul Almond’s Attempt to Generalize AI has reached part 13 (pdf), which introduces the idea of reflexive outputs.  Earlier pieces in the series and non-pdf versions are on the same site.

Gilbert Wesley Purdey has produced the T=0 Complexity Theory of Consciousness, which is based on a new attempt to tackle the surprisingly durable problem of defining consciousness, and is grounded in ‘a reformed-Macleanian overview of the evolution of the brain’. The idea that ‘Consciousness is an emergent property of the grey matter of the neo-cortex’ is one which many will accept fairly readily, but there are some unusual and interesting conclusions developed in the course of the discussion.

Carey R Carlson, whose views I briefly discussed a while ago has produced a new paper Causal Set Theory and the Origin of Mass-ratio.

Huping Hu has added a forum to the JCER site.

Is CERN, with its Large Hadron Collider, on the brink of revealing the ultimate nature of consciousness? Deepak Chopra seems to think that the Higgs boson, which CERN’s experiments might discover, has something to do with it. ‘Could this in fact be where materialism destroys itself from within?’ he asks, which seems like one of those Interesting Historical Questions To Which The Answer Is No.

My understanding of the Higgs boson is slight, but I don’t really see how Chopra’s conclusion follows here: if the existence of the Higgs boson is demonstrated, so far as I know that enables a resolution and tidying-up of some issues in physics, mainly to do with where particles get their mass from. (In my naivety I think I might have been inclined to think that mass was a metaphysical necessity on the grounds that you couldn’t make a coherent universe out of things which were under no obligation to keep still…)  That seems to be something that can only tend to strengthen materialism; but even if the boson isn’t found, we’ll merely be looking for a different account of mass. We won’t really be any worse off than we are already; arguably better off in having eliminated another blind alley; so if anything materialism is likely to be looking stronger whatever CERN may come up with.

It looks as though Chopra is really just using the Higgs as a pretext for raising again two earlier arguments. First, he thinks mere physics cannot account for the complex structures and the intentionality found in the world; we therefore need to invoke God – not in the form of an old man with a white beard, but a sort of universal consciousness.

I suppose you could connect universal consciousness with Higgs in a way. According to the Higgs theory, mass arises out of a universal Higgs field with certain interesting properties. We could hypothesise the existence of a comparable consciousness field, which endows certain entities passing through it with experience in broadly the way the Higgs field gives particles their mass. It’s not a theory I feel inclined to take up, but it might offer a way of developing an attractively clear panexperientialism, one where we might even be able to do some sums and make some predictions about the free-floating momentary experiences which I assume would be the equivalent, in this theory, of the famous boson.

But in fact that’s not quite what Chopra means: his second argument is rather that the role of the observer in collapsing wave functions in certain interpretations of quantum physics points us towards a reality in which consciousness is fundamental. He sees this as a step along the road to the perspective of Vedanta, where Brahman the all-inclusive consciousness is the self-interacting dynamic of observer, observed and process of observation.

Even on the most helpful interpretation of quantum physics, I think this reads much too much into the science: in fact I’d go so far as to say that if your reading of quantum physics requires the adoption of universal consciousness, a rather large ontological price, it seems a clear sign that your reading has gone wrong somewhere.

Generally I’d say the march of science is taking us away from universal consciousness rather than towards it. There was a time when it was reasonable to assume that consciousness was naturally out there in divine or panpsychic form and that human consciousness was best explained as some of that natural awareness taking possession of a body. Human consciousness was an inexplicable mystery and referring it to a hypothetical basic element of reality was an economical hypothesis.

Now, however, although consciousness remains mysterious, we can sort of see at least the broad outlines of the sort of way in which it might be naturalised as part of the functioning of a brain. There are some nasty gaps to say the least of it (the ‘meaning’ which Chopra alludes to can fairly be considered one), and many would still say the job is impossible, or impossible to our limited minds: but we’ve got a better hypothesis to work on and quite a bit of evidence that we’re broadly going the right way.

That partial clarification is enough to put the boot on the other foot, and leave us needing more explanation of God, or the universal consciousness. The story of human consciousness as we understand it now relies on the detailed physical interactions of biological material which is itself the product of a process of evolution: but this won’t do for divine or universal minds which have no physical structure and never competed for survival: they need some other explanation, of which there is no sign.

This need not be fatal for spiritual or panpsychic explanations: it might offer a bracing challenge and a spur to investigation and new theories. But it certainly suggests to me that waiting for quantum theory to vindicate Vedantic philosophy is, to put it mildly, optimistic.

(PS – The picture is meant to be Higgs, not Chopra.)

Does the idea of unconscious free will even make sense? Paula Droege, in the recent JCS, seems to think it might. Generally experiments like Libet’s famous ones, which seemed to show that decisions are made well before the decider is consciously aware of them, are considered fatal to free will. If the conscious activity came along only after the matter had been settled, it must surely have been powerless to affect it (there are some significant qualifications to this: Libet himself, for example, considered there was a power of last-minute veto which he called ‘free won’t’ – but still the general point is clear). If our conscious thoughts were irrelevant, it seems we didn’t have any say in the matter.

However, this view implies a narrow conception of the self in which unconscious processes are not really part of me and I only really consist of that entity that does all the talking. Yet in other contexts, notably in psychoanalysis, don’t we take the un- or sub-conscious to be more essential to our personality than the fleeting surface of consciousness, to represent more accurately what we ‘really’ want and feel? Droege, while conceding that if we take the narrow view there’s a good deal in the sceptical position, would prefer a wider view in which unconscious acts are valid examples of agency too. She would go further and bring in social influences (though it’s not entirely clear to me how the effects of social pressure can properly be transmuted into examples of my own free will), and she offers the conscious mind the consolation prize of being able to influence habits and predispositions which may in turn have a real causal influence on our actions.

I suppose there are several ways in which we exercise our agency. We perhaps tend to think of cases of conscious premeditation because they are the clearest, but in everyday life we just do stuff most of the time without thinking about it much, or very explicitly. Many of the details of our behaviour are left to ‘autopilot’, but in the great majority of cases the conscious mind would nevertheless claim these acts as its own. Did you stop at the traffic light and then move off again when it turned green? You don’t really remember doing it, but are generally ready to agree that you did. In unusual cases, we know that people sometimes even elaborate or confabulate spurious rationales for actions they didn’t really determine.

But it’s much more muddled than that. We do also at times seek to disown moral responsibility for something done when we weren’t paying proper attention, or where our rational responses were overwhelmed by a sudden torrent of emotion. Should someone who responds to the sight of a hated enemy by swerving to collide with the provoker be held responsible because the murderous act stems from emotions which are just as valid as cold calculation? Perhaps, but sometimes the opposite is taken to be the case, and the overwhelming emotion of a crime passionnel can be taken as an excuse. Then again few would accept the plea of the driver who says he shouldn’t be held responsible for an accident because he was too drunk to drive properly.

I think there may be an analogy with the responsibility held by the head of a corporation: the general rule is that the buck stops with the chief, even if the chief did not give orders for the particular action which subordinates have taken; in the same way we’re presumed by default to be responsible for what we do: but there are cases where control is genuinely and unavoidably lost, no matter what prudent precautions the chief may have put in place. There may be cases where the chief properly has full and sole responsibility; in other cases where the corporation has blundered on in pursuit of its own built-in inclinations it may be appropriate for the organization as a whole to accept blame for its corporate personality: and where confusion reigned for reasons beyond reasonable control, no responsibility may be assigned at all.

If that’s right, then Droege is on to something; but if there are two distinct grades of responsibility in play, there ought really to be two varieties of free will; the one exercised explicitly by the fully conscious me, and the other by ‘whole person’ me, in which the role of the conscious me, while perhaps not non-existent is small and perhaps mostly indirect. This is an odd thought, but if, like Droege, we broadly accept that Libet has disproved the existence of the first variety of free will, it means we don’t have the kind we can’t help believing in, but do have another kind we never previously thought of – which seems even odder.

Kristinn R. Thorisson wants artificial intelligence to build itself.  Thorisson was the creator of Gandalf*, the ‘communicative humanoid’ who was designed in a way that amply disproved Frank Zappa’s remark:

“The computer … can give you the exact mathematical design, but what’s missing is the eyebrows.”

Thorisson proposes that constructionism must give way to constructivism (pdf) if significant further progress towards artificial general intelligence is to be made. By constructionism, he means a traditional ‘divide and conquer’ approach in which the overall challenge is subdivided, modules for specific tasks are more or less hand-coded and then the results are bolted together. This kind of approach, he says, typically results in software whose scope is limited, which suffers from brittleness of performance, and which integrates poorly with other modules.  Yet we know that a key feature of general intelligence, and particularly of such features as global attention is a high level of very efficient integration, with different systems sharing heterogeneous data to produce responsive and smoothly coordinated action.

Thorisson considers some attempts to achieve better real-world performance through enhanced integration, including his own, and acknowledges that a lot has been achieved. Moreover, it is possible to extend these approaches further and achieve more: but the underlying problems remain and in some cases get worse: a large amount of work goes into producing systems which may perform impressively but lack flexibility and the capacity for ‘cognitive growth’. At best, further pursuit of this line is likely to produce improvements on a linear scale and “Even if we keep at it for centuries…  basic limitations are likely to asymptotically bring us to a grinding halt in the not-too-distant future.”

It follows that a new approach is needed and he proposes that it will be based on self-generated code and self-organising architectures. Thorisson calls this ‘constructivism’, which is perhaps not an ideal choice of name, since there are several different constructivisms in different fields already. He does not provide a detailed recipe for constructivist projects, but mentions a number of features he thinks are likely to be important. The first, interestingly, is temporal grounding – he remarks that in contrast to computational systems, time appears to be integral to the operation of all examples of natural intelligence. The second is feedback loops (but aren’t they a basic feature of every AI system?); then we have Pan-Architectural Pattern Matching, Small White-Box Components (White-Box as opposed to Black-Box, ie simple modules whose function is not hidden), and Architecture Meta-Programming and Integration.

Whether or not he’s exactly right about the way forward, Kristinsson’s criticisms of traditional approaches seem persuasive, the more so as he has been an exponent of them himself. They also raise some deeper questions which, as a practical man, he is not concerned with. One issue, indeed, is whether we’re dealing here with difficulties in practice or difficulties in principle. Is it just that building a big AGI is extremely complex, and hence in practice just beyond the scope of the resources we can reasonably expect to deploy on a traditional basis? Or is it that there is some principled problem which means that an AGI can never be built by putting together pre-designed modules?

On the face of it, it seems plausible that the problem is one of practice rather than principle, and is simply a matter of the huge complexity of the task. After all, we know that the human brain, the only example we have of successful general intelligence, is immensely complex, and that it has quirky connections between different areas. This is one occasion when Nature seems to have been indifferent to the principles of good, legible design; but perhaps ‘spaghetti code’ and a fuzzy allocation of functions is the only way this particular job can be done;  if so, it’s only to be expected that the sheer complexity of the design is going to defeat any direct attempt to build something similar.

Or we could look at it this way. Suppose constructivism succeeds, and builds a satisfactory AGI. Then we can see that in principle it was perfectly possible to build that particular AGI by hand, if only we’d been able to work out the details. Working out the details may have proved to be way beyond us, but there the thing is: there’s no magic that says it couldn’t have been put together by other methods.

Or is there? Could it be that there is something about the internal working of an AGI which requires a particular dynamic balance, or an interlocking state of several modules, that can’t be set up directly but only approached through a particular construction sequence – one that amounts to it growing itself? Is there after all a problem in principle?

I must admit I can’t see any particular reason for thinking that’s the way things are, except that if it were so, it offers an attractive naturalistic explanation of how human consciousness might be, as it were, gratuitous: not attributable to any prior design or program, and hence in one sense the furthest back we can push explanation of human thoughts and actions. If that’s true, it in turn provides a justification for our everyday assumption that we have agency and a form of free will. I can’t help finding that attractive; perhaps if the constructivist approaches Thorisson has in mind are successful this will become clearer in the next few years.

* For anyone worried about the helmet, I should explain that this Gandalf was based on a dwarf from Icelandic cosmogony, not Tolkien’s wizard of the same name.

Daniel Dennett not sceptical enough about qualia? It seems unlikely. Dennett’s trenchant view can be summed up in two words:  ‘What qualia?”.  It makes no sense, he would say, for us to talk about ineffable items of direct experience: things which by their definition, we can’t talk about. That’s not to say we can’t talk about our experience of the world: we just need to talk about it in third-person, heterophenomenological terms. Instead of claiming to discuss people’s first-person inner experience, we discuss what they report about their first-person  inner experience. In fact, if we think about it carefully, we’ll realise that’s all we could ever do, all we’ve ever done: really, whatever we may have supposed, all phenomenology is actually heterophenomenology; all discussion is necessarily in third-person, objective, scrutable, effable terms.

Typically Dennett’s is a relatively lonely voice ranged against those who would assert that qualia, direct private phenomenal experiences, are knowably, undeniably real, however hard they may be to explain and to reconcile with the objective third-person world described by science. Now Justin Sytsma (‘Dennett’s Theory of the Folk Theory’, JCS Vol 17, no 3-4) interestingly takes a different tack, suggesting that in fact Dennett has conceded too much by accepting that the folk theory, what ordinary people naively believe about their own experience, includes belief in qualia. He quotes Dennett saying:

“there seem to be qualia, because it really does seem as if science has shown us that colors can’t be out there, and hence must be in here…”

Reasonably, if somewhat unphilosophically, Sytsma treats what people actually believe as an empirical matter, something we can test; in a sense we could say that this is turning heterophenomenology on itself. It turns out, apparently, that Dennett’s assumption is false: in fact people don’t regard, say, redness, as an ineffable mental quality, but as a real property of things in the world.  Perhaps ordinary people are more sophisticated than Dennett has given them credit for: perhaps less; not sufficiently aware of  ‘what science has shown us‘ for it to have had much impact on what they believe.

What are we to make of this? Well, one issue is that there is an inbuilt tension in the entity we’re trying to discuss: Sytsma is talking about folk theories: but folk beliefs are really what we have when we have no theories: a folk theory is a kind of contradiction in terms. Julian of Norwich, I think, said that the worst thing about heretics was that they forced honest Christians to determine the truth of theological propositions which pious folk could otherwise have ignored; in a similar way we might argue that philosophical experimenters force their subjects into addressing tricky phenomenological questions which would otherwise never have troubled them.

Sytsma, then, by asking his subjects questions, was not evoking their previously-held views on phenomenology so much as engendering these views for the first time. There is an obvious danger that the terms of the question would tend to influence the form of the views evoked; but really that doesn’t matter because whatever view Sytsma evoked, it would be different to the no-view that his subjects held initially. Is the folk view pro or anti qualia?  The most accurate answer is probably ‘no’.

However,  forensically Dennett must be in the right.  If we want to establish a position, we need to argue against its negation; even if the majority or ‘the folk’ favour our view, we must instead argue against an arbitrary opponent; in fact against all the most plausible and persuasive opponents we can think of.  Even if Dennett was wrong about what people generally believe, tactically it was correct to assume that they disagreed with him: without being unduly elitist, what the majority, or the folk, or the man on the Clapham omnibus actually happen to think, is in this case philosophically uninteresting. We need not worry about whether we should endorse a sceptical theory about Dennett’s sceptical theory about folk theories. Isn’t that a relief?