Friday, 24 June 2016

AI and the (non-)Mystery of Consciousness
On Wednesday night at the Philosophers’ Cafe meeting, Patrick Conroy gave us an excellent overview of the hot and very contentious issues regarding the present status and future prospects for artificial intelligence (AI).  The complexities of the topic include definitional questions - ‘artificial,’ ’intelligence’ ’singularity’ - the present capabilities of today’s supercomputers, the startlingly ambitious predictions about AI’s future, whether a machine can become conscious, what kind of intelligence is being discussed, concerns about losing control of ‘rogue’ or ‘malevolent’ machines, and many others.
Radical opposition to the strong or true AI project from critics like me turns on the fact that vast resources - billions of dollars - are being spent on research that is decades away, if not longer, from producing the promised super-smart machines at a time when the institutions for helping people become smarter are being systematically defunded by a neoliberal business class that regards workers as disposable people.  On purely philosophical grounds, the project looks idiotic when, as many critics have pointed out, we don’t even understand human biologically based consciousness very well, much less what consciousness would be like as an “emergent property” of electronic complexity or as an upload from a brain to a central processing unit.  Are the dreams of superconscious computers at all realistic?  I think not.
The so-called Hard Problem of Consciousness is thought by many to hold the key to understanding the prospects for AI.  “How can a physical brain produce non-physical consciousness of mental events?”  So says the classic problem, with no definitive answer anywhere in sight, leading to general agreement by optimists and pessimists alike that consciousness is mysterious.  With at least one exception:  Galen Strawson, analytic philosopher and literary critic who currently holds a chair in philosophy at the University of Texas.  In a recent piece for the New York Times, Strawson makes the startling claim that consciousness “is utterly unmysterious” and that the Hard Problem has migrated somewhere else.  Let’s take a look.
Our source is:  Galen Strawson, “Consciousness Isn’t a Mystery,” New York Times, May 16, 2016.

GS says consciousness the only thing in the world we can claim to know.  “It is utterly unmysterious.”  Referencing Bertrand Russell a century or so ago, he appears to be invoking Russell’s distinction between knowledge by acquaintance and knowledge by description.    He is right, of course, in the sense that my immediate experience of anything is right there for me in the field of my awareness, effortlessly apprehended, requiring no reasoning or interpretation.  I look across the street and see a blue object and instantly recognize it as a car, but even prior to the act of recognition I am immediately aware of (acquainted with) a blue patch in my visual field.  I see it naked and exposed just as it is.  That I am seeing a blue something or other is, as Strawson says, utterly unmysterious, completely present and clear.  Similarly, if I am thinking of 5 + 3 = 8, I am  immediately aware of thinking “5 + 3 = 8.”  There is nothing mysterious about my direct experience of my own thinking.  There it is, just 5+3 = 8, pure and simple, all by itself.  I have no need of proof that I am thinking that thought, and no one can disprove it.  The same is true for any other experience:  feeling a pain in my toe or hearing a police siren, etc..  As GS says, I know these experiences immediately, because “having an experience is knowing the experience.” 

Frankly, I find the phenomenon of ‘having an experience’ pretty mysterious.  For example, who or what is this ’I’ that is having the experience? However, let’s go along with GS for a while, granting that immediate experience is unproblematic. But now he executes an astonishing right angle turn: as if by automatic transmission, he shifts the Hard Problem away from consciousness and parks it in the physics lab. By contrast with consciousness, he says, what is “deeply mysterious” is the “nature of physical stuff.” He cites Richard Feynman’s assertion that no one understands quantum theory. Hence, no one understands matter. Then GS quotes Russell for an exception: “The nature of physical stuff is mysterious “except insofar as consciousness is itself a form of physical stuff.”  A “startling statement” indeed, but what exactly does it mean?  

It seems to mean (on materialistic premises) that, whereas the objective study of matter - particles and forces in physics - continues to reveal more and more mysteries, there is a subset of physical objects not studied by physics, namely experiences, that we do know perfectly well - no mysteries, as noted above.  So when we “have conscious experiences we learn something about the intrinsic nature of (some) physical stuff, because “conscious experience is itself a form of physical stuff.”
Here is the argument:
1.  Conscious experience is a form of physical stuff.

2. When we have conscious experiences, we have direct knowledge of the intrinsic nature of those experiences.

3. Therefore, when we have conscious experiences, we have direct knowledge of the intrinsic nature of (some) physical stuff.
That's a strange argument - logically valid, yes, but we can’t help but sense that there is something seriously wrong with it.  First of all, it seems odd for GS to say that in our everyday lives, as we move about our world seeing, hearing, tasting, and doing things, we know in each and every one of those experiences exactly what matter is; we experience the ‘intrinsic nature’ of matter directly.  But if we enter a laboratory to do some physics, we are immediately confronted by a deep mystery; we do not know the intrinsic nature of matter, even though that’s what physics is supposed to tell us.  This is not a fatal objection to Strawson’s argument.  It just seems odd.
More serious is Strawson’s failure to provide any supporting argument for his first premise.  Like nearly all of his materialistic brethren, he simply asserts that consciousness is physical, because in fact it cannot be proven.  Employing the old principle that what is easily asserted is easily denied, we can comfortably reply, “Sorry, consciousness is not physical.” We could go for coffee at this point, but wait … GS says we have made a “Very Large Mistake.”
The Mistake, he says, is to think we have only two choices when confronted by the Hard Problem:  dualism or eliminativism (consciousness is unreal). In search of a third alternative, GS tries the  old ‘you can’t prove the opposite’ ploy.  He writes:
…. Many make the same mistake today — the Very Large Mistake (as Winnie-the-Pooh might put it) of thinking that we know enough about the nature of physical stuff to know that conscious experience can’t be physical. We don’t. We don’t know the intrinsic nature of physical stuff, except — Russell again — insofar as we know it simply through having a conscious experience. …. we don’t know the intrinsic nature of physical stuff in spite of all that physics tells us. In particular, we don’t know anything about the physical that gives us good reason to think that consciousness can’t be wholly physical.

This looks suspiciously like the fallacy of reasoning from ignorance.  But even if it’s not, it’s absurd to claim that, although we don’t know the intrinsic nature of matter (as opposed to its surface appearances), there is nothing in what we do know (from physics) about matter that gives us any reason to think that consciousness can’t be physical.  That’s like saying that although we don’t know the intrinsic nature of justice, what we do know gives us no grounds for judging rape to be unjust.
Of course physics gives us no reason for thinking consciousness can’t be physical.  Physics is about the physical.  It has nothing whatsoever to say about what may or may not be non-physical.  That’s a question for philosophers, not physicists, and as a philosopher GS is simply wrong:  we have plenty of reasons for thinking consciousness is not physical, but those come from philosophy of science, not physics.  All we need is some basic understanding of consciousness and of the structure of scientific method, both of which we have in plentiful supply, as GS himself  knows.
Having written about this at length in earlier posts  (October 9 and 20, 2015), I will content myself here with just two arguments.  First, consciousness is subjective, physics is objective.  Studies of physical matter and studies of consciousness are carried out from two entirely different perspectives, one that looks outward toward objects that are publicly observable; the other looks inward to a domain that is entirely private, unavailable to direct access by anyone else - the world of my sensations, thoughts, imaginings, and feelings.
The subjective cannot be reduced to the objective, if only for the simple fact that in order to carry out the reduction, the philosopher has to use his own “I think…” which is his consciousness and which cannot be made both subject and object at the same time.  He could claim coherently, if absurdly, that other people’s consciousness is nothing but brain events, but he cannot make the same claim about his own without committing a performative contradiction.  The act of saying, “I am just a sequence of brain events” exposes its own absurdity instantly.
Second, the “rules and equations” of physics are useless when aimed at my experience of tasting garlic or the twinge of jealousy I feel when my girl friend spends too much time talking with another guy.  Those are real experiences, not abstractions.
Nor are the biological concepts of neuroscience any help.  Biological concepts refer to entities like cells, which have mass, and energy transfers between brain cells that can be measured by instruments.  The taste of garlic on my tongue or my thought of the square root of 2 have neither mass nor energy nor any other physical characteristic that can be observed or measured.  What has no physical characteristics cannot be physical.  Therefore, consciousness is not physical.
With his current line of thinking, GS thinks he has laid the Hard Problem of Consciousness to rest.  I sympathize with his wish to do so, but the path he has chosen leads nowhere.  His argument that we do not know the intrinsic nature of physical stuff except when we are having conscious experiences, if not entirely nonsensical, looks like a sophistical trick to recast the Hard Problem of Consciousness as something other than it is.  The reasoning seems to work only because it begs the question by assuming what the argument is supposed to prove.
Speaking of the Hard Problem of Consciousness, let’s take a quick look at what it is and why it causes such headaches among philosophers.  It is usually stated:  “How does the brain produce consciousness?  How does a physical, biological organ give rise to non-physical conscious experiences?”  When a philosophical problem resists solution for a very long time, one begins to suspect there might be something wrong with the question.  As I have argued in earlier posts (October 9 and 20, 2015), there is indeed something very wrong with the Hard Problem, namely its underlying assumption that the brain produces or gives rise to consciousness.  Far from being self-evident, that belief is really an article of faith, an a priori assumption that just has to be true, because materialistic philosophers really, really want it to be true.  But what if it’s not true?  After all, no one has ever seen a brain produce a thought, and no one ever will for the aforementioned reasons.
What if ‘produce’ is a flawed metaphor for the relationship between the brain and the mind?  If that is the case, then not even a million years of philosophizing will reveal how brain-producing-mind takes place.  So where does that leave us?  We have the honest confession from from physicists themselves that physical matter is intrinsically mysterious.  We also know that the Hard Problem has survived Prof. Strawson’s deconstruction effort.  
Looks like Mystery is Us - still.                                                                                                                                             


  1. I would think thoughts are weighty to their degree like physical things: if I think the wrong thing, that can directly lead to doing the wrong thing. E.g.: If I have the experience of seeing a crime committed, it's a big deal if I don't take the action to report it. What happened in my mind when seeing the crime--the understanding that it's illegal--carries weight. I think it's true that you're privy to the intrinsic nature of that--seeing the main issue, in other words.

    1. Thoughts can indeed have a kind of 'weightiness,' as in the example you mention. You seem to mean that thoughts can have serious consequences in the physical world, hence 'weighty.' True. We often use physical metaphors to characterize thoughts and other mental activities we perform. Thoughts or questions can be 'deep,' as can feelings. A mood can be 'dark' or 'bright.' A book, which is a medium of thoughts, may be described as 'light' or 'heavy' reading. These are useful metaphors, but we do not mean them literally. We do not weigh a book to see whether it is light or heavy reading. A crime caused by a thought, say a theft of $50, can be observed and perhaps even measured in various ways, but thinking about committing the crime, the thought itself ("I'm going to steal $50.") cannot be weighed or measured. In themselves, thoughts have no physical properties.

    2. A few decades ago I viewed a TV show on a study that was done: a small animal (dog or cat) was placed on a scale just prior to its death. After it passed they saw it had actually lost weight--ounces or pounds, I don't recall how much.

    3. The conclusion drawn was I believe that the animal's spirit or soul had weight.

  2. This comment has been removed by the author.