attentionAre our minds being dumbed by digits – or set free by unreading?

Frank Furedi notes  that it has become common to deplore a growing tendency to inattention. In fact, he says, this kind of complaint goes back to the eighteenth century. Very early on the failure to concentrate was treated as a moral failing rather than simple inability; Furedi links this with the idea that attention to proper authority is regarded as a duty, so that inattention amounts to disobedience or disrespect. What has changed more recently, he suggests, is that while inattention was originally regarded as an exceptional problem, it is now seen as our normal state, inevitable: an attitude that can lead to fatalism.

The advent of digital technology has surely affected our view. Since the turn of the century or earlier there have been warnings that constant use of computers, and especially of the Internet, would change the way our brains worked; would damage us intellectually if not morally. Various kinds of damage have been foreseen; shortened attention span, lack of memory, dependence on images, lack of concentration, failure of analytical skills and inability to pull the torrent of snippets into meaningful structures. ‘Digital natives’ might be fluent in social media and habituated to their own strange new world, but there was a price to pay. The emergence of Homo Zappiens has been presented as cause for concern, not celebration.

Equally there have been those who suggest that the warnings are overstated. It would, they say, actually be strange and somewhat disappointing if study habits remained exactly the same after the advent of an instant, universal reference tool; the brain would not be the highly plastic entity we know it to be if it didn’t change its behaviour when presented with the deep interactivity that computers offer; and really it’s time we stopped being surprised that changes in the behaviour of the mind show up as detectable physical changes in the brain.

In many respects, moreover, people are still the same, aren’t they? Nothing much has changed. More undergraduates than ever cope with what is still a pretty traditional education. Young people have not started to find the literature of the dark ages before the 1980s incomprehensible, have they? We may feel at times that contemporary films are dumbed down, but don’t we remember some outstandingly witless stuff from the 1970s and earlier? Furedi seems to doubt that all is well; in fact, he says, undergraduate courses are changing, and are under pressure to change more to accommodate the flighty habits of modern youth who apparently cannot be expected to read whole books. Academics are being urged to pre-digest their courses into sets of easy snippets.

Moreover, a very respectable recent survey of research found that some of the alleged negative effects are well evidenced.

 Growing up with Internet technologies, “Digital Natives” gravitate toward “shallow” information processing behaviors characterized by rapid attention shifting and reduced deliberations. They engage in increased multitasking behaviors that are linked to increased distractibility and poor executive control abilities. Digital natives also exhibit higher prevalence of Internet-related addictive behaviors that reflect altered reward-processing and self-control mechanisms.

So what are we to make of it all? Myself, I take the long view; not just looking back to the early 1700s but also glancing back several thousand years. The human brain has reshaped its modus operandi several times through the arrival of symbols and languages, but the most notable revolution  was surely the advent of reading. Our brains have not had time to evolve special capacities for the fairly recent skill of reading, yet it has become almost universal, regarded as a natural accomplishment almost as natural as walking. It is taken for granted in modern cities – which increasingly is where we all live – that everyone can read. Surely this achievement required a corresponding change in our ability to concentrate?

We are by nature inattentive animals; like all primates we cannot rest easy – as a well-fed lion would do – but have to keep looking for new stimuli to feed our oversized brains. Learning to read, though, and truly absorbing a text, requires steady focus on an essentially linear development of ideas. Now some will point out that even with a large tome, we can skip; inattentive modern students may highlight only the odd significant passage for re-reading as though Plato need really only have written fifty sentences; some courteously self-effacing modern authors tell you which chapters of their work you can ignore if you’re already expert on A, or don’t like formulae, or are only really interested in B. True; but to me those are just the exceptions that highlight the existence of the rule that proper  books require concentration.

No wonder then, that inattention first started to be seriously stigmatised in the eighteenth century, just when we were beginning to get serious about literacy; the same period when a whole new population of literate women became the readership that made the modern novel viable.

Might it not be that what is happening now is that new technology is simply returning us to our natural fidgety state, freed from the discipline of the long, fixed text? Now we can find whatever nugget of information we want without trawling through thousands of words; we can follow eccentric tracks through the intellectual realm like an excitable dog looking for rabbits. This may have its downside, but it has some promising upsides too: we save a lot of time, and we stand a far better chance of pulling together echoes and correspondences from unconnected matters, a kind of synergy which may sometimes be highly productive. Even those old lengthy tomes are now far more easily accessible than they ever were before. The truth is, we hardly know yet where instant unlimited access and high levels of interactivity will take us; but perhaps unreading, shedding some old habits, will be more a liberation than a limitation.

But now I have hit a thousand words, so I’d better shut up.

boxers…for two theories?

Ihtio kindly drew my attention to an interesting paper which sets integrated information theory (IIT) against its own preferred set of ideas – semantic pointer competition (SPC). I’m not quite sure where this ‘one on one’ approach to theoretical discussion comes from. Perhaps the authors see IIT as gaining ground to the extent that any other theory must now take it on directly. The effect is rather of a single bout from some giant knock-out tournament of theories of consciousness (I would totally go for that, incidentally; set it up, somebody!).

We sort of know about IIT by now, but what is SPC? The authors of the paper, Paul Thagard and Terrence C Stewart, suggest that:

consciousness is a neural process resulting from three mechanisms: representation by firing patterns in neural populations, binding of representations into more complex representations called semantic pointers, and competition among semantic pointers to capture the most important aspects of an organism’s current state.

I like the sound of this, and from the start it looks like a contender. My main problem with IIT is that, as was suggested last time, it seems easy enough to imagine that a whole lot of information could be integrated but remain uniluminated by consciousness; it feels as if there needs to be some other functional element; but if we supply that element it looks as if it will end up doing most of the interesting work and relegate the integration process to something secondary or even less important. SPC looks to be foregrounding the kind of process we really need.

The authors provide three basic hypotheses on which SPC rests;

H1. Consciousness is a brain process resulting from neural mechanisms.
H2. The crucial mechanisms for consciousness are: representation by patterns of firing in neural populations, binding of these representations into semantic pointers, and competition among semantic pointers.
H3. Qualitative experiences result from the competition won by semantic pointers that unpack into neural representations of sensory, motor, emotional, and verbal activity.

The particular mention of the brain in H1 is no accident. The authors stress that they are offering a theory of how brains work. Perhaps one day we’ll find aliens or robots who manage some form of consciousness without needing brains, but for now we’re just doing the stuff we know about. “…a theory of consciousness should not be expected to apply to all possible conscious entities.”

Well, actually, I’d sort of like it to – otherwise it raises questions about whether it really is consciousness itself we’re explaining. The real point here, I think, is meant to be a criticism of IIT, namely that it is so entirely substrate-neutral that it happily assigns consciousness to anything that is sufficiently filled with integrated information. Thagard and Stewart want to distance themselves from that, claiming it as a merit of their theory that it only offers consciousness to brains. I sympathise with that to a degree, but if it were me I’d take a slightly different line, resting on the actual functional features they describe rather than simple braininess. The substrate does have to be capable of doing certain things, but there’s no need to assume that only neurons could conceivably do them.

The idea of binding representations into ‘semantic pointers’ is intriguing and seems like the right kind of way to be going; what bothers me most here is how we get the representations in the first place. Not much attention is given to this in the current paper: Thagard and Stewart say neurons that interact with the world and with each other become “tuned” to regularities in the environment. That’s OK, but not really enough. It can’t be that mere interaction is enough, or everything would be a prolific representation of everything around it; but picking out the right “regularities” is a non-trivial task, arguably the real essence of representation.

Competition is the way particular pointers get selected to enter consciousness, according to H2; I’m not exactly sure how that works and I have doubts about whether open competition will do the job. One remarkable thing about consciousness is its coherence and direction, and unregulated competition seems unlikely to produce that, any more than a crowd of people struggling for access to a microphone would produce a fluent monologue. We can imagine that a requirement for coherence is built in, but the mechanism that judges coherence turns out to be rather important and rather difficult to explain.

So does SPC deliver? H3 claims that it gives rise to qualitative experience: the paper splits the issue into two questions: first, why are there all these different experiences, and second, why is there any experience at all? On the first, the answers are fairly good, but not particularly novel or surprising; a diverse range of sensory inputs and patterns of neural firing naturally give rise to a diversity of experience. On the second question, the real Hard Problem, we don’t really get anywhere; it’s suggested that actual experience is an emergent property of the three processes of consciousness. Maybe it is, but that doesn’t really explain it. I can’t seriously criticise Thagard and Stewart because no-one has really done any better with this; but I don’t see that SPC has a particular edge over IIT in this respect either.

Not that their claim to superiority rests on qualia; in fact they bring a range of arguments to suggest that SPC is better at explaining various normal features of consciousness. These vary in strength, in my opinion. First feature up is  how consciousness starts and stops. SPC has a good account, but I think IIT could do a reasonable job, too. The second feature is how consciousness shifts, and this seems a far stronger case; pointers naturally lend themselves better to thus than the gradual shifts you would at first sight expect from a mass of integrated information. Next we have a claim that SPC is better at explaining the different kinds or grades of consciousness that fifteen organisms presumably have. I suppose the natural assumption, given IIT, would be that you either have enough integration for consciousness or you don’t. Finally, it’s claimed that SPC is the winner when it comes to explaining the curious unity/disunity of consciousness. Clearly SPC has some built-in tools for binding, and the authors suggest that competition provides a natural source of fragmentation. They contrast this with Tononi’s concept of quantity of consciousness, an idea they disparage as meaningless in the face of the mental diversity of the organisms in the world.

As I say, I find some of these points stronger than others, but on the whole I think the broad claim that SPC gives a better picture is well founded. To me it seems the advantages of SPC mainly flow from putting representation and pointers at the centre. The dynamic quality this provides, and the spark of intentionality, make it better equipped to explain mental functions than the more austere apparatus of IIT. To me SPC is like a vehicle that needs overhauling and some additional components (some of those not readily available); it doesn’t run just now but you can sort of see how it would. IIT is more like an elegant sculptural form which doesn’t seem to have a place for the wheels.

humoursWorse than wrong? A trenchant piece from Michael Graziano likens many theories of consciousness to the medieval theory of humours; in particular the view that laziness is due to a build up of phlegm. It’s not that the theory is wrong, he says – though it is – it’s that it doesn’t even explain anything.

To be fair I think the theory of the humours was a little more complex than that, and there is at least some kind of hand-waving explanatory connection between the heaviness of phlegm and slowness of response. According to Graziano such theories flatter our intuitions; they offer a vague analogy which feels metaphorically sort of right – but, on examination, no real mechanism. His general point is surely very sound; there are indeed too many theories about conscious experience that describe a reasonably plausible process without ever quite explaining how the process magically gives rise to actual feeling, to the ineffable phenomenology.

As an example, Graziano mentions a theory that neural oscillations are responsible for consciousness; I think he has in mind the view espoused by Francis Crick and others that oscillations at 40 hertz give rise to awareness. This idea was immensely popular at one time and people did talk about “40 hertz” as though it was a magic key. Of course it would have been legitimate to present this as an enigmatic empirical finding, but the claim seemed to be that it was an answer rather than an additional question. So far as I know Graziano is right to say that no-one ever offered a clear view as to why 40 hertz had this exceptional property, rather than 30 or 50, or for that matter why co-ordinated oscillation at any frequency should generate consciousness. It is sort of plausible that harmonising on a given frequency might make parts of the brain work together in some ways, and people sometimes took the view that synchronised firing might, for example, help explain the binding problem – the question of how inputs from different senses arriving at different times give rise to a smooth and flawlessly co-ordinated experience. Still, at best working in harmony might explain some features of experience: it’s hard to see how in itself it could provide any explanation of the origin or essential nature of consciousness. It just isn’t the right kind of thing.

As a second example Graziano boldly denounces theories based on integrated information. Yes, consciousness is certainly going to require the integration of a lot of information, but that seems to be a necessary, not a sufficient condition. Intuitively we sort of imagine a computer getting larger and more complex until, somehow, it wakes up. But why would integrating any amount of information suddenly change its inward nature? Graziano notes that some would say dim sparks of awareness are everywhere, so that linking them gives us progressively brighter arrays. That, however, is no explanation, just an even worse example of phlegm.

So how does Graziano explain consciousness? He concedes that he too has no brilliant resolution of the central mystery. He proposes instead a project which asks, not why we have subjective experience, but why we think we do: why we say we do with such conviction. The answer, he suggests, is in metacognition. (This idea will not be new to readers who are acquainted with Scott Bakker’s Blind Brain Theory.) The mind makes models of the world and models of itself, and it is these inaccurate models and the information we generate from them that makes us see something magic about experience. In the brief account here I’m not really sure Graziano succeeds in making this seem more clear-cut than the theories he denounces. I suppose the parallel existence of reality and a mental model of reality might plausibly give rise to an impression that there is something in our experience over and above simple knowledge of the world; but I’m left a little nervous about whether that isn’t another example of the kind of intuition-flattering the other theories provide.

This kind of metacognitive theory tends naturally to be a sceptical theory; our conviction that we have subjective experience proceeds from an error or a defective model, so the natural conclusion, on grounds of parsimony if no others, is that we are mistaken and there is really nothing special about our brain’s data processing after all.

That may be the natural conclusion, but in other respects it’s hard to accept. It’s easy to believe that we might be mistaken about what we’re experiencing, but can we doubt that we’re having an experience of some kind? We seem to run into quasi-Cartesian difficulties.

Be that as it may Graziano deserves a round of applause for his bold (but not bilious) denunciation of the phlegm.

backwardIn the course of this review of Carlo Rovelli’s Seven Brief Lessons on Physics, Alva Noë asks a fascinating question: does it make any sense to imagine that one might think backward?

In physics we are accustomed to thinking that entropy always increases; things run down, spread out, differences are evened out and available energy always decreases. Are cognitive processes like that, always going in one underlying direction? Are they like a river which flows to the sea, or are they more like a path we can take in either direction?

Well, we can’t talk backwards without careful preparation, and some kinds of conscious thought resemble talking to oneself. In fact, we can’t perform most tasks backwards without rehearsal. Unless one takes the trouble to learn zedabetical order, we cannot easily recite our letters in reverse. So it doesn’t seem we can do that kind of backward thinking, at least. We can, of course, read a written sentence in the reverse word order and I suppose that in that sense we can think the same sentence backwards, or ‘backwards sentence same the think’; but we might well doubt whether that truly reverses the thought.

On the other hand, on another level, there seems to be no inherent direction to a chain of thought. If I think of an egg, it makes me think of breakfast, and breakfast makes me think of getting up; but thinking about getting up might equally have prompted thoughts of breakfast, which in turn might have brought an egg to mind. The association of ideas goes both ways. Logical thought needs a little more care. ‘If p then q’ does not allow us to conclude that if q then p – but we can say that not-q gives us not-p. With due attention to the details there doesn’t seem to be a built-in direction to deduction.

The universal presence of memory in all our thoughts seems a deal-breaker on reversibility, though. We cannot forget things at will; memories constantly accumulate; and we cannot deliberately unthink a thought in such a way as to erase it from recollection. This one-way effect seems to be equally true of certain kinds of thought on another level. We have, let’s say, a problem in mind; all the clues we need are known. At some point they come together to point to the right conclusion – Aha! Unless we find a flaw in our reasoning, the meshing of the clues cannot be unpicked. We can’t ununderstand; the process of comprehension seems as irrevocable as the breaking of an eggshell or the cooking of a omelette.

There’s an odd thing about it though. The breaking of the egg is an example of the wider principle of entropy; the shell is destroyed, and later the protein is denatured by heat. The solving of the problem, by contrast, is constructive; we’re left with more than we had before and our mental contents are better structured, not worse. Learning and understanding seems like a process of growth. Like the growth of a plant it is of course just a local reversal of entropy, which has to be paid for through the use of a lot of irretrievable energy; still, it’s remarkable (as is the growth of a plant, after all).

Hang on, though. Isn’t it the case that we might have been entertaining a dozen different ideas about the possible solution to our problem?  Once we have the answer those other hypotheses are dropped and indeed may even be truly forgotten. More than that, the right answer may let us simplify our ideas, the way Copernicus let us do away with epicycles and all the rest of the Ptolemaic system nobody has to think about any more. Occam tells us to minimise the number of angels we require for our theory, so isn’t the growth of understanding sometimes a synthesis which actually has the character of a reductive simplification? That isn’t usually a reversal as such, but doesn’t it involve in some sense the unthinking of certain thoughts? David Lodge somewhere has a character feel a pang of pity for all the Marxist professors in the universities of no-longer-communist countries who must presumably unspool years of patient theoretical exegesis in order to start understanding the world again.

Well, yes, but I don’t think that is truly an unthinking or reverse cogitation. Is it perhaps more like a plant managing to grow downwards? So no, as Noë implied in the first place, it doesn’t really make sense to imagine we might think backward.

observationAre we being watched? Over at Aeon, George Musser asks whether some AI could quietly become conscious without our realising it. After all, it might not feel the need to stop whatever it was doing and announce itself. If it thought about the matter at all, it might think it was prudent to remain unobserved. It might have another form of consciousness, not readily recognisable to us. For that matter, we might not be readily recognisable to it, so that perhaps it would seem to itself to be alone in a solipsistic universe, with no need to try to communicate with anyone.

There have been various scenarios about this kind of thing in the past which I think we can dismiss without too much hesitation. I don’t think the internet is going to attain self-awareness because however complex it may become, its simply isn’t organised in the right kind of way. I don’t think any conventional digital computer is going to become conscious either, for similar reasons.

I think consciousness is basically an expanded development of the faculty of recognition. Animals have gradually evolved the ability to recognises very complex extended stimuli; in the case of human beings things have gone a massive leap further so that we can recognise abstractions and generalities. This makes a qualitative change because we are no longer reliant on what is coming in through our sense from the immediate environment; we can think about anything, even imaginary or nonsensical things.

I think this kind of recognition has an open-ended quality which means it can’t be directly written into a functional system; you can’t just code it up or design the mechanism. So no machines have been really good candidates; until recently. These days I think some AI systems are moving into a space where they learn for themselves in a way which may be supported by their form and the algorithms that back them up, but which does have some of the open-ended qualities of real cognition. My perception is that we’re still a long way from any artificial entity growing into consciousness; but it’s no longer a possibility which can be dismissed without consideration; so a good time for George to be asking the question.

How would it happen? I think we have to imagine that a very advanced AI system has been set to deal with a very complex problem. The system begins to evolve approaches which yield results and it turns out that conscious thought – the kind of detachment from immediate inputs I referred to above – is essential. Bit by bit (ha) the system moves towards it.

I would not absolutely rule out something like that; but I think it is extremely unlikely that the researchers would fail to notice what was happening.

First, I doubt whether there can be forms of consciousness which are unrecognisable to us. If I’m right consciousness is a kind of function which yields purposeful output behaviour, and purposefulness implies intelligibility. We would just be able to see what it was up to. Some qualifications to this conclusion are necessary. We’ve already had chess AIs that play certain end-games in ways that don’t make much sense to human observers, even chess masters, and look like random flailing. We might get some patterns of behaviour like that. But the chess ‘flailing’ leads reliably to mate, which ultimately is surely noticeable. Another point to bear in mind is that our consciousness was shaped by evolution, and by the competition for food, safety, and reproduction. The supposed AI would have evolved in consciousness in response to completely different imperatives, which might well make some qualitative difference. The thoughts of the AI might not look quite like human cognition.  Nevertheless I still think the intentionality of the AI’s outputs could not help but be recognisable. In fact the researchers who set the thing up would presumably have the advantage of knowing the goals which had been set.

Second, we are really strongly predisposed to recognising minds. Meaningless whistling of the wind sounds like voices to us; random marks look like faces; anything that is vaguely humanoid in form or moves around like a sentient creature is quickly anthropomorphised by us and awarded an imaginary personality. We are far more likely to attribute personhood to a dumb robot than dumbness to one with true consciousness. So I don’t think it is particularly likely that a conscious entity could evolve without our knowing it and keep a covert, wary eye on us. It’s much more likely to be the other way around: that the new consciousness doesn’t notice us at first.

I still think in practice that that’s a long way off; but perhaps the time to think seriously about robot rights and morality has come.

bokkenThe latest JCS features a piece by Christopher Curtis Sensei about the experience of achieving mastery in Aikido. It seems he spent fifteen years cutting bokken (an exercise with wooden swords, don’t ask me), becoming very proficient technically but never satisfying the old Sensei. Finally he despaired and stopped trying; at which point, of course, he made the required breakthrough. He needed to stop thinking about it. You do feel that his teacher could perhaps have saved him a few years if he had just said so explicitly – but of course you cannot achieve the state of not thinking about something directly and deliberately. Intending to stop thinking about a pink hippo involves thinking about a pink hippo; you have to do something else altogether.

This unreflective state of mind crops up in many places; it has something to do with the desirable state of ‘flow’ in which people are said to give their best sporting or artistic performances; it seems to me to be related to the popular notion of mindfulness, and it recalls Taoist and other mystical ideas about cleared minds and going with the stream. To me it evokes Julian Jaynes, who believed that in earlier times human consciousness manifested itself to people as divine voices; what we’re after here is getting the gods to shut up at last.

Clearly this special state of mind is a form of consciousness (we don’t pass out when we achieve it) and in fact on one level I think it is very simple. It’s just the absence of second-order consciousness, of thoughts about thoughts, in other words. Some have suggested that second-order thought is the distinctive or even the constitutive feature of human consciousness; but it seems clear to me that we can in fact do without it for extended periods.

All pretty simple then. In fact we might even be able to define it physiologically – it could be the state in which the cortex stops interfering and let’s the cerebellum and other older bits of the brain do their stuff uninterrupted – we might develop a way of temporarily zapping or inhibiting cortical activity so we can all become masters of whatever we’re doing at the flick of a switch. What’s all the fuss about?

Except that arguably none of the foregoing is actually about this special state of mind at all. What we’re talking about is unconsidered thought, and I cannot report it or even refer to it without considering it; so what have I really been discussing? Some strange ghostly proxy? Nothing? Or are these worries just obfuscatory playing with words?
There’s another mental thing we shouldn’t, logically, be able to talk about – qualia. Qualia, the ineffable subjective aspect of things, are additional to the scientific account and so have no causal powers; they cannot therefore ever have caused any of the words uttered or written about them. Is there a link here? I think so. I think qualia are pure first-order experiences; we cannot talk about them because to talk about them is to move on to second-order cognition and so to slide away from the very thing we meant to address. We could say that qualia are the experiential equivalent of the pure activity which Curtis Sensei achieved when he finally cut bokken the right way. Fifteen years and I’ll understand qualia; I just won’t be able to tell anyone about it…

red lampWas Libet wrong? The question has often been asked since the famous experiments which found that a detectable “Readiness Potential” (RP) showed that our decisions were made half a second before they entered our consciousness. There are plenty of counter arguments but the findings themselves have been reproduced and seem unassailable.

However, Christian Jarrett reports two new pieces of research which shed fresh light on the issue (Jarrett’s book Great Myths of the Brain is very sensible, btw, and should be read by all science journalists).

The first piece of research  shows that although an action may be prepared by the brain before we know we’ve decided to act, we can still change our minds. Subjects were asked to press a button at a moment of their choice after a green light came on. However, if a red light appeared before they pressed, they were asked to refrain. The experimenters then detected the RP which showed the subjects were about to press the button, and used the red light to try to cancel the intention. They found that although there was a ‘point of no return’, so long as the red light appeared in time the subjects were able to hold off and not press the button after all.

That means that a significant qualification has to be added to Libet’s initial findings – but it’s one Libet himself had already come up with. He was aware that the emergent action could be suppressed, and memorably said it showed that while we might not have free will, we could still have ‘free won’t’. I don’t know much use that is. In the experiment the subjects simply responded to a red light, but if the veto is to make consciousness effective again we seem to need the veto decision to happen instantly and overtake the action decision somehow. That seems problematic if not paradoxical. I also get quite confused trying to imagine what it would be like to veto mentally a decision you’re not yet aware of having made. Still, I suppose we must be grateful for whatever residual degree of freedom the experiments allow us.

The second piece of research  calls into question the nature of the RP itself. Libet’s research more or less took it for granted that the RP was a reliable sign that an action was on the way, but the new findings suggest that it is really just part of the background ebb and flow of neural noise. Actions do arise when the activity crosses a certain threshold, but the brain is much quicker about getting to that level when the background activity is already high.

That certainly adds some complexity to the picture, but I don’t think it really dispels the puzzle of Libet’s results. The RP may be fuzzier than we thought and it may not have as rigid a link to action as we thought – but it’s still possible to predict the act before we’re aware of the decision. What if we redesigned the first experiment? We tell the subject to click on a target at any time after it appears; but we detect the RP and whip the target away a moment before the click every time. The subject can never  succeed. The fact that that is perfectly possible surely remains more than a little unsettling.

Go boardThe recent victory scored by the AlphaGo computer system over a professional Go player might be more important than it seems.

At first sight it seems like another milestone on a pretty well-mapped road; significant but not unexpected. We’ve been watching games gradually yield to computers for many years; chess notoriously, was one they once said was permanently out of the reach of the machines. All right, Go is a little bit special. It’s an extremely elegant game; from some of the simplest equipment and rules imaginable it produces a strategic challenge of mind-bending complexity, and one whose combinatorial vastness seems to laugh scornfully at Moore’s Law – maybe you should come back when you’ve got quantum computing, dude! But we always knew that that kind of confidence rested on shaky foundations; maybe Go is in some sense the final challenge, but sensible people were always betting on its being cracked one day.

The thing is, Go has not been beaten in quite the same way as chess. At one time it seemed to be an interesting question as to whether chess would be beaten by intelligence – a really good algorithm that sort of embodied some real understanding of chess – or by brute force; computers that were so fast and so powerful they could analyse chess positions exhaustively. That was a bit of an oversimplification, but I think it’s fair to say that in the end brute force was the major factor. Computers can play chess well, but they do it by exploiting their own strengths, not by doing it through human-style understanding. In a way that is disappointing because it means the successful systems don’t really tell us anything new.

Go, by contrast, has apparently been cracked by deep learning, the technique that seems to be entering a kind of high summer of success. Oversimplifying again, we could say that the history of AI has seen a contest between two tribes; those who simply want to write programs that do what’s needed, and those that want the computer to work it out for itself, maybe using networks and reinforcement methods that broadly resemble the things the human brain seems to do. Neither side, frankly, has altogether delivered on its promises and what we might loosely call the machine learning people have faced accusations that even when their systems work, we don’t know how and so can’t consider them reliable.

What seems to have happened recently is that we have got better at deploying several different approaches effectively in concert. In the past people have sometimes tried to play golf with only one club, essentially using a single kind of algorithm which was good at one kind of task. The new Go system, by contrast, uses five different components carefully chosen for the task they were to perform; and instead of having good habits derived from the practice and insights of human Go masters built in, it learns for itself, playing through thousands of games.

This approach takes things up to a new level of sophistication and clearly it is yielding remarkable success; but it’s also doing it in a way which I think is vastly more interesting and promising than anything done by Deep Thought or Watson. Let’s not exaggerate here, but this kind of machine learning looks just a bit more like actual thought. Claims are being made that it could one day yield consciousness; usually, if we’re honest, claims like that on behalf of some new system or approach can be dismissed because on examination the approach is just palpably not the kind of thing that could ever deliver human-style cognition; I don’t say deep learning is the answer, but for once, I don’t think it can be dismissed.

Demis Hassabis, who led the successful Google DeepMind project, is happy to take an optimistic view; in fact he suggests that the best way to solve the deep problems of physics and life may be to build a deep-thinking machine clever enough to solve them for us (where have I heard that idea before?). The snag with that is that old objection; the computer may be able to solve the problems, but we won’t know how and may not be able to validate its findings. In the modern world science is ultimately validated in the agora; rival ideas argue it out and the ones with the best evidence wins the day. There are already some emergent problems, with proofs achieved by an exhaustive consideration of cases by computation that no human brain can ever properly validate.

More nightmarish still the computer might go on to understand things we’re not capable of understanding. Or seem to: how could we be sure?

Alfred SmeeCharles Babbage was not the only Victorian to devise a thinking machine.

He is, of course, considered the father, or perhaps the grandfather, of digital computing. He devised two remarkable calculating machines; the Difference Engine was meant to produce error-free mathematical tables for navigation or other uses; the Analytical Engine, an extraordinary leap of the imagination, would have been the first true general-purpose computer. Although Babbage failed to complete the building of the first, and the second never got beyond the conceptual stage, his achievement is rightly regarded as a landmark, and the Analytical Engine routines published by Lady Lovelace in 1843 with a translation of Menabrea’s description of the Engine, have gained her recognition as the world’s first computer programmer.

The digital computer, alas, went no further until Turing a hundred years later; but in 1851 Alfred Smee published The Process of Thought adapted to Words and Language together with a description of the Relational and Differential Machines – two more designs for cognitive mechanisms.

Smee held the unusual post of Surgeon to the Bank of England – in practice he acted as a general scientific and technical adviser. His father had been Chief Accountant to the Bank and little Alfred had literally grown up in the Bank, living inside its City complex. Apparently, once the Bank’s doors had shut for the night, the family rarely went to the trouble of getting them unlocked again to venture out; it must have been a strangely cloistered life. Like Babbage and other Victorians involved in London’s lively intellectual life, Smee took an interest in a wide range of topics in science and engineering, with his work in electro-metallurgy leading to the invention of a successful battery; he was a leading ophthalmologist and among many other projects he also wrote a popular book about the garden he created in Wallington, south of London – perhaps compensating for his stonily citified childhood?

Smee was a Fellow of the Royal Society, as was Babbage, and the two men were certainly acquainted (Babbage, a sociable man, knew everyone anyway and was on friendly terms with all the leading scientists of the day; he even managed to get Darwin, who hated socialising and suffered persistent stomach problems, out to some of his parties). However, it doesn’t seem the two ever discussed computing, and Smee’s book never mentions Babbage.

That might be in part because Smee came at the idea of a robot mind from a different, biological angle. As a surgeon he was interested in the nervous system and was a proponent of ‘electro-biology’, advocating the modern view that the mind depends on the activity of the brain. At public lectures he exhibited his own ‘injections’ of the brain, revealing the complexity of its structure; but Golgi’s groundbreaking methods of staining neural tissue were still in the future, and Smee therefore knew nothing about neurons.

Smee nevertheless had a relatively modern conception of the nervous system. He conducted many experiments himself (he used so many stray cats that a local lady was moved to write him a letter warning him to keep clear of hers) and convinced himself that photovoltaic effects in the eye generated small currents which were transmitted bio-electrically along the nerves and processed in the brain. Activity in particular combinations of ‘nervous fibrils’ gave rise to awareness of particular objects. The gist is perhaps conveyed in the definition of consciousness he offered in an earlier work, Principles of the Human Mind Deuced from Physical Laws:

When an image is produced by an action upon the external senses, the actions on the organs of sense concur with the actions in the brain; and the image is then a Reality.
When an image occurs to the mind without a corresponding simultaneous action of the body, it is called a Thought.
The power to distinguish between a thought and a reality, is called Consciousness.

This is not very different in broad terms from a lot of current thinking.

In The Process of Thought Smee takes much of this for granted and moves on to consider how the brain deals with language. The key idea is that the brain encodes things into a pyramidal classification hierarchy. Smee begins with an analysis of grammar, faintly Chomskyan in spirit if not in content or level of innovation. He then moves on rapidly to the construction of words. If his pyramidal structure is symbolically populated with the alphabet different combinations of nervous activity will trigger different combinations of letters and so produce words and sentences. This seems to miss out some essential linguistic level, leaving the impression that all language is spelled out alphabetically, which can hardly be what Smee believed.

When not dealing specifically with language the letters in Smee’s system correspond to qualities and this pyramid stands for a universal categorisation of things in which any object can be represented as a combination of properties. (This rather recalls Bishop Wilkins’ proposed universal language, in which each successive letter of each noun identifies a position in an hierarchical classification, so that the name is an encoded description of the thing named.)

At least, I think that’s the way it works. The book goes on to give an account of induction, deduction, and the laws of thought; alas, Smee seems unaware of the problem described by Hume and does not address it. Instead, in essence, he just describes the processes; although he frames the discussion in terms of his pyramidal classification his account of induction (he suggests six different kinds) comes down to saying that if we observe two characteristics constantly together we assume a link. Why do we do that – and why is it we actually often don’t? Worse than that, he mentions simple arithmetic (one plus one equals two, two times two is four) and says:

These instances are so familiar we are apt to forget that they are inductions…

Alas, they’re not inductions. (You could arrive at them by induction, but no-one ever actually does and our belief in them does not rest on induction.)

I’m afraid Smee’s laws of thought also stand on a false premise; he says that the number of ideas denoted by symbols is finite, though too large for a man to comprehend. This is false. He might have been prompted to avoid the error if he had used numbers instead of letters for his pyramid – because each integer represents an idea; the list of integers goes on forever, yet our numbering system provides a unique symbol for every one? So neither the list of ideas nor the list of symbols can be finite. Of course that barely scratches the surface of the infinitude of ideas and symbols, but it helps suggest just how unmanageable a categorisation of every possible idea really is.

But now we come to the machines designed to implement these processes. Smee believed that his pyramidal structure could be implemented in a hinged physical mechanism where opening would mean the presence or existence of the entity or quality and closing would mean its absence. One of these structures provides the Relational Machine. It can test membership of categories, or the possession of a particular attribute, and can encode an assertion, allowing us to test consistency of that assertion with a new datum. I have to confess to having only a vague understanding of how this would really work. He allows for partial closing and I think the idea is that something like predicate calculus could be worked out this way. He says at one point that arithmetic could be done with this mechanism and that anyone who understands logarithms will readily see how; I’m afraid I can only guess what he had in mind.

It isn’t necessary to have two Relational Machines to deal with multiple assertions because we can take them in sequence; however the Differential Machine provides the capacity to compare directly, so that we can load into one side all the laws and principles that should guide a case while uploading the facts into the other.

Smee had a number of different ideas about how the machines could be implemented, and says he had a number of part-completed prototypes of partial examples on his workbench. Unlike Babbage’s designs, his were never meant to be capable of full realisation, though; although he thinks it is finite he says the Relational Machine would cover London and the mechanical stresses would destroy it immediately if it were ever used; moreover, using it to crank out elementary deductions would be so slow and tedious people would soon revert to using their wonderfully compact and efficient brains instead. But partial examples will helpfully illustrate the process of thought and help eliminate mistakes and ‘quibbles’. Later chapters of the book explore things that can go wrong in legal cases, and describe a lot of the quibbles Smee presumably hopes his work might banish.

I think part of the reason Smee’s account isn’t clearer (to me, anyway) is that his ideas were never critiqued by colleagues and he never got near enough to a working prototype to experience the practical issues sufficiently. He must have been a somewhat lonely innovator in his lab in the Bank and in fairness the general modernity of his outlook makes us forget how far ahead of his time he was. When he published his description of his machines, Wundt, generally regarded as the founder of scientific psychology, was still an undergraduate. To a first approximation, nobody knew anything about psychology or neurology. Logic was still essentially in the long Aristotelian twilight – and of course we know where computing stood. It is genuinely remarkable that Smee managed, over a hundred and fifty years ago, to achieve a proto-modern, if flawed, understanding of the brain and how it thinks. Optimists will think that shows how clever he was; pessimists will think it shows how little our basic conceptual thinking has been updated by the progress of cognitive science.

minskyMarvin Minsky, who died on Sunday, was a legend.  Here’s a slightly edited version of my 2004 post about him.

Linespace

 

Is it time to rehabilitate Marvin Minsky? As a matter of fact, I don’t think he was ever dishabilitated (so to speak) but it does seem to be generally felt that there are a couple of black marks against his name. The most widely-mentioned count against him and his views is a charge of flagrant over-optimism about the prospects for artificial intelligence. A story which gets quoted over and over again has it that he was so confident about duplicating human cognitive faculties, even way back in the 1970s when the available computing power was still relatively modest, that he gave the task of producing a working vision system to one of his graduate students as a project to sort out over the summer.

The story is apocryphal, but one can see why it gets repeated so much. The AI sceptics like it for obvious reasons, and the believers use it to say “I may seem like a gung-ho over-the-top enthusiast, but really my views are quite middle-of-the-road, compared to some people. Look at Marvin Minsky, for example, who once…”

Still, there is no doubt that Minsky did at one time predict much more rapid progress than has, in the event materialized: in 1977 he declared that the problem of creating artificial intelligence would be substantially solved within a generation.

The other and perhaps more serious charge against him is that in 1969, together with Seymour Papert, he gave an unduly negative evaluation of Frank Rosenblatt’s ‘Perceptron’ (an early kind of neural network device which was able to tackle simple tasks such as shape recognition). Their condemnation, based on the single-layer version of the perceptron rather than more complex models, is considered to have led to the effective collapse of Rosenblatt’s research project and a long period of eclipse for networks before a new wave of connectionist research came along and claimed Rosenblatt as an unfairly neglected forerunner.

There’s something in both these charges, but surely in fairness neither ought to be all that damaging? Optimism can be a virtue, without which many long and difficult enterprises could not get started, and Minsky’s was really no more starry-eyed than many others. The suggestion of AI within a generation does no more at most than echo Turing’s earlier forecast of human-style performance by the end of the century, and although it didn’t come true, you would have to be a dark pessimist to deny that there were (and perhaps still are) some encouraging signs.

It seems to be true that Minsky and Papert, by focusing on the single-layer perceptron alone, did give an unduly negative view of Rosenblatt’s ideas – but if researchers were jailed for giving dismissive accounts of their rivals’ work, there wouldn’t be many on the loose today. The question is why Minsky and Papert’s view had such a strong negative influence when a judicious audience would surely have taken a more balanced view.

I suspect that both Minsky’s optimism and his attack on the perceptron should properly be seen as crystallizing in a particularly articulate and trenchant form views which were actually widespread at the time: Minsky was not so much a lonely but influential voice as the most conspicuous and effective exponent of the emergent consensus.

What then, about his own work? I take the most complete expression of his views to be “The Society of Mind”. This is an unusual book in several ways – for one thing it is formatted like no other book I have ever seen, with each page having its own unique heading. It has an upbeat tone, compared to many books in the consciousness field, which tend to be written as much against a particular point of view as for one. It is full of thought-provoking points, and is hard to read quickly or to summarise adequately, not because it is badly or obscurely written (quite the contrary) but because it inspires interesting trains of thought which take time to mull over adequately.

The basic idea is that each simple task is controlled by an agent, a kind of sub-routine. A set of agents which happen to be active when a good result is achieved get linked together by a k-line. The multi-level hierarchical structures which gradually get built up allow complex and highly conditional forms of behaviour to be displayed. Building with blocks is used as an example, starting with simple actions such as picking up a block, and gradually ascending to the point where we debate whether to go on playing a complex block-building game or go off to have lunch instead. Ultimately all mental activity is conducted by structures of this kind.

This is recognizably the way well-designed computer programs work, and it also bears a plausible resemblance to the way we do many things without thinking (when we move our arm we don’t think about muscle groups, but somehow somewhere they do get dealt with); but it isn’t a very intuitively appealing general model of human thought from an inside point of view. It naturally raises some difficulties about how to ensure that appropriate patterns of behaviour can be developed in novel circumstances. There are many problems which arise if we just leave the agents to slug it out amongst themselves – and large parts of the book are taken up with the interesting solutions Minsky has to offer. The real problem (as always) arises when we want to move out of the toy block world and deal with the howling gale of complexity presented by the real world.

Minsky’s solution is frames (often compared with Roger Schank’s similar strategy of scripts). We deal with reality through common sense, and common sense is, in essence, a series of sets of default assumptions about given events and circumstances. When we go to a birthday party, we have expectations about presents, guests, hosts, cakes and so on which give us a repertoire of appropriate to deploy and a context in which to respond to unfolding events. Alas, we know that common sense has so far proved harder to systematize than expected – so much so that these days the word ‘frame’ in a paper on consciousness is highly likely to be part of the dread phrase ‘frame problem’.

The idea that mental activity is constituted by a society of agents who themselves are not especially intelligent is an influential one, and Minsky’s version of it is well-developed and characteristically trenchant. He has no truck at all with the idea of a central self, which in his eyes is pretty much the same thing as an homunculus, a little man inside your head. Free will, for him, is a delusion which we are unfortunately stuck with. This sceptical attitude certainly cuts out a lot of difficulties, though the net result is perhaps that the theory deals better with unconscious processes than conscious ones. I think the path set out by Minsky stops short of a real solution to the problem of consciousness and probably cannot be extended without some unimaginable new development. That doesn’t mean it isn’t a worthwhile exercise to stroll along it, however.