The Narcissus Syndrome

narcissusThe Loebner prize, for the program best able to simulate a human being in conversation, along the lines of the celebrated Turing Test, has been won for the second year running by Rollo Carpenter’s Jabberwacky, this time using ‘Joan’ a female personality. Last year’s contest was mentioned here . Jabberwacky seems to be making steady progress: some media attention has been attracted recently by the visual people-simulations which have been commissioned to accompany the verbal output. George, last year’s winning personality is now visible  here.

You can have a conversation with Jabberwacky yourself here : my impression is that unless you deliberately set out to trap the program, it delivers quite long stretches of plausible, if rather evasive, dialogue. I suppose the evasiveness is an inevitable product of the computer not really understanding what you’re talking about – it’s like someone trying to bluff their way through a conversation about a book that they haven’t actually read. It might disappear if the software were dealing with a limited domain, an area which it could ‘know about’ in more detail. That suggests that practical usefulness in some such restricted context (perhaps answering routine queries about a particular product, or being an interactive ‘tour guide’ in a museum) is no longer an unrealistic aspiration – though “really” passing the Turing Test in free conversation still is.

Should we be pleased or worried? Chatbot programs do not pretend to reproduce all the ineffable properties of real human thought and consciousness: they just aim to deliver good outputs by whatever means works. It has been suggested, however, that their resemblance to a full-fledged conscious being could have a subtly malign impact on our attitudes. The problem is that some people – many people, probably – are more than willing to attribute personhood to programs which haven’t any claim to it. Daniel Dennett has remarked, in connection with his own erstwhile involvement with the Loebner, that the attitude of the interlocutor was often a more important factor than the quality of the chatbot. Some sceptical people devise cunning questions which rely on knowledge of the world, or the context of the dialogue, in ways which throw even the most sophisticated and well-prepared program: but many others accept almost any grammatical output as a human-like response. It seems there is something seductive about having our own image reflected in a machine; a kind of Narcissus effect which makes us fascinated with a dialogue in which we are really the only players.

This willingness to be deceived became evident with Joseph Weizenbaum’s famous Eliza, the mother of all chatbots: he famously found his secretary conducting a long and deeply engaged conversation with the program, much to his horror. I saw the same effect on a smaller scale myself many years ago, in the days when an 80 by 25 display of glowing green characters was still the standard PC display technology. I had a simple program which drew a face on the screen using ascii characters: at random intervals it would raise an eyebrow, swivel its square eyes, smile, blink, and so on. Colleagues I had previously regarded as sensible took this thing to be far more sophisticated than it was: not human, obviously, but “maybe up to about the level of a tortoise” (it looked a bit chelonian).

We might well find, then, that if we start dealing with plausible chat-bots on a regular basis, we shall automatically start to think of them in much the same way as we think of human beings. But there are some obvious dangers in confusing people and simple machines. On the one hand, it might lead us to trust the advice of machines rather more than we should. We are already a bit prone to this, following the instructions of our in-car GPS navigation system even when it conflicts with common sense, and attributing a spurious authority to job evaluation or personality test programs which merely reflect back at us in a digested form the views we fed into them in the first place.

More seriously, we might find our instincts being tutored towards treating people the way we treat machines – as tools to be manipulated and used without any ethical significance. I think you could make a case that this sort of thing has already begun to happen: attitudes to euthanasia have certainly shifted a long way in recent times for example (if the thing’s worn out, junk it), and utilitarian calculations about patients as generators of “quality of life” are far more overtly applied in medical contexts than would once have been the case. Moreover, it seems to be more readily accepted these days that moral responsibility is largely an illusion and that economic and social conditions determine the behaviour of criminals and heroes equally. Books and other works of art merely express the writer’s place in the societal matrix rather than anything individual and ineffable. Does all this represent a welcome clarity about human nature, or a depressingly impoverished view of the world?

I understand the pessimistic view, which I think lies behind some people’s distaste for the whole idea of AI. But in the end I think it under-rates the subtlety of people’s attitudes. The fact is we are already used to a world in which all sorts of things which lack a human brain are treated in varying degrees as animate. Children do and don’t believe that their toys have live personalities; the ancient Romans and many others believed in gods that were sort of people, and sort of mere embodiments of abstract qualities. In Japan, Shinto grants the status of Kami (god, spirit, ensouled thing) to all the most salient features of nature, without people becoming confused or losing their sense of human specialness. For that matter, humanoid robots have been a part of our culture for a long time now – the word itself is over eighty years old. Perhaps Weizenbaum’s concern about his secretary was unnecessary: she may have regarded the friendly counsellor in the computer as no more real than the elusive Hearts players that help some Windows users pass the time these days. And… could it be possible that my old colleagues were to some extent just winding me up? I doubt if they would have gone back into a burning building to save tortoise-face, in any case. All in all, I think we can cope.

Of course, if we want to be really optimistic, there is another possibility – that dealing with chatbots every day will actually sharpen and improve our sense of the special qualities of real people…

Tandem triumphans

HonderichBlandulaTed Honderich , as he promised a while ago, has returned in triumph to the periodical which woundingly rejected him a couple of years ago. Back then, Honderich’s paper on ‘Consciousness as Existence‘ failed the peer review for the Journal of Consciousness Studies: now, a whole issue is devoted to a target paper from him and eleven responses.

In the interval, Honderich has not been idle. He published “On Consciousness”, a more substantial adumbration of the same theory, and others, notably Manzotti and Tonneau as mentioned here, have rallied to the banner of Radical Externalism which he has raised. There’s really no such thing as final vindication for a philosopher, but this surely represents a remarkable improvement in the reception accorded to his views, and it must be highly gratifying: all the more so because, on the whole, the reception accorded his paper is a fairly friendly, positive one.

Linespace

BitbucketIt’s a good time for Honderich – he’s on television in the UK this week presenting his views on the justification for terrorism. He is generally perceived as saying that it is justified, but of course it’s more complicated than that – I think it’s something along the lines of his cautiously asserting the possibility of a denial of the doctrine that there can in principle be no circumstance in which no less than one of the proffered justifications for terrorist acts can be seen on moral grounds to possess the qualifying characteristics mandated of it (whatever and in whichever theoretical context those may, adequately described no doubt in the relevant papers, be presumed to be) under those theories, not here to be elucidated but sufficiently indicated, perhaps, by such.

Perhaps that’s a slightly unfair parody, but I find Honderich’s expositions always suffer from a strange kind of fogginess. He seems to be asserting something very bold, but amongst all the meandering prose you get a certain impression of the fist not quite connecting. It was just the same, many years ago, with his well-received book, Punishment: it looked as if he thought all punishment was wrong, but nothing half so bold as that was ever quite asserted.

And now, consciousness: he seems to be saying, in fact he surely does say, that my consciousness of something just amounts to the thing existing, which is certainly bold: bold to the point of barminess – but it turns out that it actually amounts to the thing ‘in a way’ existing: three little words which actually stand for a hefty additional apparatus. Tim Crane, in his response, accuses Honderich of equivocating on this, and I find it hard not to agree: in some places Honderich seems to lay stress on the simplicity of the thesis that consciousness is existence, just existence: in others, he explains that what he has in mind is a new and puzzling mode of existence in a second or possibly a third metaphysical world with complicated relations of dependency on other worlds to sustain it.

Linespace

BlandulaI find Honderich’s style rather engaging: it has a kind of gently self-deprecating humour. You’re a bit unfair, aren’t you? On the one hand you accuse him of being “foggy”, and then you complain that his brief statement of the main point is too terse: doesn’t include all the details, and therefore conflicts or equivocates with what he says elsewhere.

As Honderich fairly says, anyway, this particular piece isn’t a detailed, definitive exposition, just a presentation of the theory: namely, that for you to be conscious of something is just for that thing to exist, not in the underlying physical world, nor in the form of firing neurons, but in another, perceptual world, dependent both on the physics and the neurons but separate from them. In other words, for you to see something, the actual atoms and molecules of it have to be there in the physical world, and your neurons have to be firing in whatever way they do when you notice something: but your consciousness does not consist in wither of those things: it consists in the object’s being there in your perceptual world – not inside your skull.
You may find this a ‘hefty apparatus’, but I think it has a lot of intuitive appeal. Think of what people say when they talk about their perceptions – they use phrases like “there’s something there”, or “it’s gone”, even when they are not talking about an object but about the mere appearance itself.

Linespace

BitbucketAh, mere appearances! Honderich can’t actually cope with those, can he? Since he says that your awareness of something is equivalent to it’s existing (in a way, blah de blah), it follows that illusions, dreams, imaginary things and mirages, must all also exist (in a way, yacketty smacketty). Harold Brown illustrates the difficulty quite neatly by asking about Kanisza triangles (those ones which ‘appear’ when three ‘pac-men’ are positioned so as to define their corners, but have no actual outlines). Do they exist (in a manner, yada yada)?

Honderich concedes that it isn’t clear how these triangles fit into his conceptual framework, but he reckons it is a minor issue of the kind a friendly graduate student could sort out for him. Surely it’s worse than that?

After all, one of the things Honderich particularly wants his theory to do is to banish sense-data style theories from the landscape. Many people have thought that we do not perceive the world directly, but through internal representations of some kind, and it seems to follow that all we ever really perceive is those representations. The main argument put forward for such views is the argument from error or illusion: if we perceived things directly, how could we ever be wrong about them? The error must creep in through our internal representation being different from reality. Honderich is very keen to deny this: indeed I think it is a major part of his motivation here. He asserts that perception is not perception of some representation or sense-data inside our head, it is the external existence of the thing perceived. But if you want to refute a theory whose main argument is based on problems over errors, your theory surely has to have a robust way of dealing with such errors?

Linespace

BlandulaI agree there’s more to be done in clearing up issues like this, but Honderich has a number of options open to him. He denies that conscious awareness is the perception of internal representations, but that doesn’t commit him to denying that there is any such thing as internal representations. It might well be perfectly reasonable to deny that the objects of perception are in the head while affirming that the objects of dreams and illusions are. This might imply that dreams and illusions are not perceptions, but what’s wrong with that – I’d say they’d better not be! In responding to Paul Snowdon Honderich exploits a distinction between affective and perceptual consciousness, which seems a viable enough path to take.

Linespace

BitbucketOK. I suppose it’s true that you can lash together a solution to any difficulty if you don’t care about the additional overhead in complexity and ontological commitments that you incur. As a matter of fact, the basic ontological housekeeping of Radical Externalism is its weakest point if you ask me.

I mean, you’ve got this object of perception – let’s be unimaginative and say it’s a chair. It’s not in the world of physics, it’s not in my neurons, it’s in this other place, this world of perception. Is that my world of perception, or the world of perception?

Let’s assume to begin with that there’s only one world of perception: is the chair I perceive there the same as the one you perceive? Strange if so, because it doesn’t look quite the same to you as it does to me. These differences are not really errors, just differences in point of view and the like: so Honderich can’t deploy whatever he may eventually come up with as a solution for errors (and if he could I think he’d find that his treatment of errors gradually eroded the rest of his theory away altogether). I think we’re forced to conclude that the objects of our perception are different. This looks worryingly as if the chair itself might split into two, but Honderich can retain the identity of the single chair intact in the world of physics and just allow it to have, as it were, different avatars in the world of perception.
However, if all the objects of my perception are separate from all the objects of yours, we might just as well say that we each have our own separate world of perception, containing the objects of perception special to us individually. So let’s move on to the hypothesis that we each have our own world of perception with its own objects: our own are directly accessible to us, but not at all to other people (or we should all have a perfect kind of telepathy). Now you can call such worlds external if you like, but it means nothing: if they’re particular to us and denied to everyone else, wouldn’t it be more reasonable to describe them as internal?

Linespace

BlandulaI think the point is not about perceptions being specific to individuals, but about their being distinct from my personal neuronal activity. If we can agree that they’re outside the skull, I think Honderich might not care all that much about your wanting to call them internal in some other vague sense. And your version of internality does strike me as pretty loose: my shoes are particular to me, but that doesn’t make them internal. Don’t get me wrong on all this – I’m not saying I’m signing up for membership of the Radical Externalists. But I would go along with what some of the respondents say: anything that cracks open the traditional ways we look at these things: gives us a new set of categories and concepts, must surely be welcome. Even if you don’t like the externalism, you’ve surely got to give two cheers for the radicalism?

Seeing Red

Nicholas Humphrey

Nicholas Humphrey recently produced a postscript to his book “Seeing Red”. He remarks with mild regret that some readers perhaps will still not “get it”: but if his theory is true, that’s only to be expected. If he’s right, the mysterious nature of qualia is really the point; if they were easy to understand, they wouldn’t do their job properly – a tantalising suggestion.

Actually I suspect that the most troublesome point in Humphrey’s theory, the one which most people tend to baulk at and the bit that’s hardest to “get” is his view of sensation, which he distinguishes from mere perception (the former involves having a subjective experience, the latter is just a matter of affectless acquisition of facts). The problem is that Humphrey insists that sensation “has something of the character of a bodily action”: it isn’t just passive reception. When we see red, the sensation arises from us “redding”, or to take a slightly more persuasive example, when we feel a pain, it’s because we’re hurting.

I must say that in the case of vision this seems completely counter-intuitive. Seeing colour feels very much like passive reception to me, but Humphries is very clear about its active character, referring to it as an example of agency. Agency surely requires control, the ability to act or refrain from acting; yet colour vision does not appear to be voluntary, like other examples of agency – I can’t switch to monochrome vision in the same way I can decide to shut my eyes. It would be a shame if this put people off the rest of the theory, however, because it seems to me that the argument works just as well if we take “redding” to be a purely reactive affair, devoid of real agency.

What is “redding” and other similar actions, anyway? Although Humphrey doesn’t describe it this way, we could regard it as a kind of second-order perception: a special sort of response to our own immediate reaction to stimuli. Once upon a time, our primitive single-celled ancestors might have responded to red light with a particular twitch of the membrane: as their descendants became more complex, Humphrey suggests, the initial reaction might have been disconnected from the twitch and used instead as the input to slightly more developed decision-making processes. Using your own reaction to stimuli as a source of information about the world has its limits, however, and ultimately our somewhat nearer ancestors would develop proper sensory apparatus separately. They would then have two channels providing information about the world: one a straightforward channel of perception, the other an indirect source based on monitoring the descendant of the old twitch-for-red reaction, still firing away somewhere in our brain. It’s this internal monitoring of the phantom twitch which provides the qualia; but in fact it does more than that.

Humphrey points out that although strictly speaking the present moment is an instant of zero duration, we do not experience it that way. Our experience is of a short stretch of time, with recent events still to some degree in our thoughts. He refers to this as the “thick present”, and suggests that a likely mechanism is a feedback circuit which causes present impressions to go on reverberating in our minds for a short time until they die away. It’s a plausible idea, though I suspect it may be a bit more complex than that: memory must surely be involved, and I would imagine that the level of attention we devote to things strongly affects the extent of their persistence and perhaps even the length of the perceived “thick present”. Be that as it may, the real point is that Humphrey thinks his “redding” mechanism provides just the right kind of feedback loop to cause the kind of reverberations he is proposing.

So far so good, but why should monitoring our own vestigial twitches be anything like subjective experience? Why isn’t it just like monitoring our own vestigial twitches? The fail-safe objection to any theory of qualia is the one that goes “Yes, but I can imagine all that stuff you describe happening in my brain or wherever, and me still not having any qualia”. For once, though, there is at least a tentative answer.

Humphrey arrives at this answer by asking himself a different question. Why have we retained this complicated feedback mechanism, he asks. Evolution tends to weed out redundant devices, so there must be a strong presumption that a qualia-generating facility has some very definite survival value. Perhaps we can spot what this is by considering the case of blindsight, the strange phenomenon of people who can’t see consciously, but can still point accurately to dots on a screen, or pick up objects without fumbling. Humphrey has a long acquaintance with blindsight: in fact he described a case in an experimental monkey before the human version had been documented. He tells the poignant story of a woman whose sight defect was corrected late in life. Because her brain had had no visual inputs it had not developed the processing arangements needed to deal with them, so although her eyes now worked perfectly she remained functionally blind. Or so it seemed: Humphrey suspected, and then confirmed by experiment, that although she had no conscious experience of vision, she did have a kind of general blindsight which, if she chose to use it, allowed her to perform all kinds of practical tasks which blindness had previously rendered impossible.

The strange thing is, she seemed unimpressed and unengaged by this new ability: did not enjoy using it, and eventually chose to give up on it and revert to being the blind person which, in her subjective experience, she had always been. Why? Qualia, Humphrey suggests, are profoundly engaging: they make things seem to matter. Without them, his subject was unable to get herself to take any real interest in her blindsighted abilities; and without them we should all have an anaemic grey existence composed only of dull information. It seems to follow that philosophical zombies, people just like us but without qualia, would turn out to have a zombie-like lack of interest in life too, and hence be indifferent competitors in the battle for survival.

But why do qualia make things seem to matter? Because, in the final analysis, they are part of us; whereas our objective senses tell us about the external world, qualia are really our own internal reactions: surely it is only logical that we should find them more vivid and more important than simple facts about the world. In fact, they are also the basis of our sense of self, and help explain our devotion to the survival of that self, too.

We can now see why Humphrey half-expects people not to “get it”. If it were clear to us that qualia were merely mental reactions, their valuable motivational effect would be dissipated, as would the useful delusion that we have a remarkable, immaterial self. It is only to be expected that evolution will have made the truth difficult to grasp, in our own interests.

This is a clever, cogent and rather ingenious argument, but I think I can see some rational reasons to doubt its truth. While perceiving qualia as part of ourselves might logically make them more important to us, I’m not sure that the idea of an immaterial self naturally encourages us to take a keen interest in the survival of our bodies. In fact, we know quite well that the opposite is the case: people who believe strongly in the immortality of the soul are less concerned about death, and may even seek out martyrdom.

More generally, does the wonderfulness of qualia really motivate us to be better survivors? It seems as likely to have us wasting time marvelling at the beauty of the daffodils while a tiger creeps up from behind: in survival terms it seems hard to beat the value of those grey but accurate facts. Actually, it is usually taken to be an essential property of qualia that they have no effect on our behaviour – if you’re talking about something that changes the decisions we make, on this view, you’re not talking about qualia at all.

I certainly have some reservations, in any case, about drawing many conclusions from blindsight examples. It seems quite likely to me that paying continuous attention to the subliminal or unconscious influences through which blindsight presumably works might be an exhausting, difficult, and uncertain business, one which the unfortunate woman in the case Humphrey quotes might easily have found burdensome, rather than simply unengaging through its lack of qualia.

There’s something a little fishy, too, about the argument that qualia have a special impact because they are part of us. I can only care particularly in this way about things I know to be part of me – yet my redding has its special effect just because I don’t recognise that it is internal (in fact, I imagine it to be a feature of red objects out there).
So although Humphrey’s argument is a clever, interesting, and indeed quite persuasive one, I’m not quite convinced. Perhaps it’s true that in one way or another evolution has destined me for disbelief in this case.

(More here, about Humphrey’s 2011 book Soul Dust. And a better picture – I think my drawing skills have come on a bit.)