Monday, December 24, 2012

How To Build A Free-Will Machine

One of the hottest topics in physics and philosophy these days is the question of free will. Do we humans make truly free choices in the world, or is this impression merely an illusion? It certainly seems as if we have free will, but if science teaches us anything, it’s that we can’t always trust our intuition: Feathers respond to gravity the same as bowling balls (in a vacuum), light does not travel infinitely fast, and the surface of a pond is not flat, but curves slightly with the shape of the Earth. Is free will wrong, too?

Those who subscribe to the reductionist school of thought would say that we do not have free will. According to this argument, a complete understanding of the world can be reduced to particles strictly obeying the laws of physics; therefore, whatever physical state you (i.e., the particles of your brain/body) and your environment were in prior to making a choice, there’s one and only one corresponding state afterward, and that means only one choice — predetermined by the earlier state of the atoms and molecules in your body. Some take this to the extreme, arguing that the future of everything in the universe is already decided, and that this future could be predicted with perfect accuracy in principle (if not in practice), given complete knowledge* of the universe’s present state. A different line of argument points to cognition experiments: It is now well known that our brain has already settled on a decision about a half-second before we consciously sense that we are deciding; therefore our conscious sensation of free will must be an illusion, it’s argued.

The term I use for this kind of thinking is “retarded.” I mean no offense; I use the term literally, that these arguments are regressive and backward, unimaginative, crippled by the Einstellung effect — they are based on old paradigms and leave no room for new, “outside the box” ways of thinking. (We used to think patterns on the surface of Mars were canals built by Martians; after all, we humans build canals on Earth, right? And, the Martians have human faces. That’s retarded thinking.) The experimental argument against free will is retarded, because it assumes that only our conscious self is capable of making free decisions. What if the actual free decision happens a half-second earlier in the subconscious, and only the conscious aspect of free will (“I think I’ll make a left turn here”) is the illusory part? The high-level executive functions of the brain, which include thoughts and sensations, are only a small part of consciousness, like the images displayed on a computer screen. There’s a lot more going on at deeper levels than it seems, and this is where free will may reside.

That leaves the physical, reductionist argument against free will. Prominent scientists including physicist Paul Davies and mathematician George Ellis reject this as well, on the grounds that strict reductionism does not apply to living systems. In the science literature there has been an explosion of research and theory on the role of information in biological systems, and we are seeing a groundswell of acknowledgment that in living organisms, information plays a causal role on the atoms and molecules of life. This recognition of “top-down” effects is changing our view of the bottom-up mechanisms which, according to traditional reductionist thinking, drive everything in the universe.

If information is fundamental to the way organisms (such as humans) operate, can we demonstrate that free will really exists by describing it in terms of information, rather than atoms and molecules? How would that work?

Let’s consider one of the most human-brain-like machines in the world, the Jeopardy-playing IBM computer Watson. Watson uses a sophisticated statistical approach: Given a Jeopardy clue, Watson compares keywords and strings of words with a vast database of information, runs a slew of algorithms simultaneously, and then comes up with a list of possible responses, assigning each a confidence level. If the confidence level of one response is sufficiently high, Watson rings in and gives a response. Google Translate and Apple’s Siri use similar statistical approaches. But no one in their right mind would say that Watson or Siri has free will; given exactly the same prompt and the same database — what scientists call initial conditions — the result will be entirely predictable. Despite being an incredibly complex computer, Watson is still not as complex as the simplest one-celled animal, let alone a human brain. So, how could we modify Watson so that it would start to exhibit qualities of free will?

Dynamics. That’s the key difference between Watson and living cells. Watson uses more or less a fixed database and operating rules, which is why, given the same clue, Watson would respond the same. But dynamics — in the form of highly complex, interacting internal changes — are one of the most obvious hallmarks of living organisms. If Watson were built with interacting dynamics, the results would be chaotic enough that Watson would begin to exhibit free-will-like qualities. For example, a random number generator could alter all statistical calculations slightly over time. That alone would make its responses more unpredictable. Another number generator could randomly remove access to sectors of the database, mimicking the imperfection of biological memory and recall. Watson’s thresholds — the risks it is willing to take — could go up and down slightly with time, as well as in response to external conditions (how far into the game it is, Watson’s score against those of its competitors, even the instantaneous temperature and air pressure). Watson could be given “moods”: If it missed a couple of clues in a row, it might get “bummed out” and avoid risks for a while. There could be positive and negative feedback mechanisms that either exaggerate or reduce risk-taking over time, based on several of the other factors. Changes in light levels and noises (such as a burst of laughter or applause) could “distract” Watson, causing the confidence levels to dip or fluctuate uncontrollably, with some distractions being longer than others, based on factors such as recent performance and the scores. “Fatigue” could set in, with the threshold for distraction going down not only steadily by time but also as a function of Watson’s performance and even the time of day. Watson might be given a mechanical “body” that must cooperate in order to play the game, this interplay dynamically affected by the body’s own complex dynamics and feedback mechanisms and distractions. (Too much ringing in? “Hand” cramps up.) And so on.

Given all of these extra dynamics, would Watson have human-type free will? Not quite. That would require piling on astronomical layers of complexity. Phrases in clues might conjure specific “memories” from its “life” that could either help or hurt performance; it could have multiple competing internal influences or “dialogues,” akin to Freud’s id and superego (or like Gollum/Sméagol from Lord of the Rings); and it could have advanced “emotions” such as jealousy or contempt, those emotions modulated by its “memories” as well as every other factor I’ve mentioned. That doesn’t even touch on the decidedly human skill of analytically understanding (and misunderstanding!) the true meanings of the clues the first place.

Regardless, adding only five or ten interacting dynamic parameters to the existing Watson would create a system sufficiently chaotic that its behavior might exhibit the free will of, say, a flatworm. Simple living creatures have enough dynamic complexity going on that it’s effectively impossible for us to recreate the same initial conditions, both internal and external, of any given choice it might have to make. So, even though a flatworm or a modified Watson will usually respond to a certain stimulus in a certain way, you can never know enough about the system to say for sure. Throw in the indeterminate/random nature of quantum-mechanical influences at the sub-cellular level (analogous to the number generators in modified Watson), and it becomes impossible even in theory to predict how choices will be made.

As far as I’m concerned, that means free will, even for a flatworm, or for a Watson. For a human being, with all of its complexities and frailties, it isn’t even a matter of debate.



* The “knower,” being a part of the universe, would need complete instantaneous knowledge of itself, including knowledge of the state of having learned the last fact about itself. Or, it would need to be external to the universe, which is defined as all that exists. Both options are logically impossible.

11 comments:

  1. I just came across your blog -- awesome. I'm listing you in my favorites. If you haven't seen it already, brother Netwriter has a great video on Free Will: http://youtu.be/-SFYER6JjuM

    ReplyDelete
  2. Interesting view, though I think the definition of free will - as applied to humans - implicit in this article puts the bar too high for the argument against it, to the point that it defies what is conventionally understood as "free will". In my view, you have free will to the extent that the decisions and choices you make are attributable to "you" (which is inescapably a metaphysical concept, as it requires you to answer the question "who/what am I?"), as opposed to your environment, genetic predispositions and personal history. Of course, there are elements of your environment, genes and personal history that could be integrated into the way you define who "you" are. For instance, if my genes make me a sociopath, and I consider sociopathy as a part of "who I am" rather than as something external to myself, then my decision to kill someone in cold blood will appear to be a "free" decision. If I don't, then the opposite will seem to be true. I see no rational or scientific way to determine what constitutes the self with any precision.
    As much as I hate metaphysics, I'm afraid there's no way to give a good answer to the question of free will without them.
    Here's an interesting article by one of my favorite legal scholars that applies the question of free will to the realm of economics: http://www.becker-posner-blog.com/2012/10/luck-wealth-and-implications-for-policy-posner.html

    ReplyDelete
  3. congratulations on making a good defence of free will in the face of current reductionist fashions....like you I think instead of an absence of free will there is in fact much more than we allow...as with its corollary consciousness...perhaps that is a gud definition of life processes..the creation and perfection of free will machines

    ReplyDelete
  4. Eddie, Love your satire stuff. But couldn't agree with you on this one.

    "it isn't even a matter of debate." - The clue to your error might be that it is in fact debated.

    "Let's consider one of the most human-brain-like machines in the world, the Jeopardy-playing IBM computer Watson." - Another error. This may give us confidence that Watson does not have free will; but you miss some points:

    Two entities, Human(H) and Watson(W), where H is assumed to have free will and W not. Show that W does not have free will, and that piling on masses of complexity still doesn't get W free will. Therefore H has free will? All W shows is that it's hard to make things that have free will, without addressing at all the matter of whether there is free will, anywhere, in anything, at all, including H. So Watson is irrelevant to the Human free will question. Except that it contributes a little more counter evidence to there being free will.

    If anything the argument should go something like this: No matter what we have tried, no matter how complex we make machines like W, they never exhibit what we would call free will. It seems rather odd, and convenient, that only H has free will. Shouldn't we re-think what free will is? And question again whether H has it or not? This is why it is a matter of debate.

    ...

    ReplyDelete
  5. ...

    The ONLY 'evidence' we have that humans have free will is that to humans it feels like it - and then, only subjectively by the individual. We can't tell from watching others that they have free will at all. We infer they do, based on the 'theory of mind' by which humans project onto other humans what they are feeling themselves: if I think I have free will, and he seems to behave like me, and report that he has free will, then I guess we both have. This is hopelessly inadequate non-science. And there is ample evidence that when humans think they are acting freely they are not - not least priming experiments, whereby subjects are set-up to make a specific choice, and then later they rationalise about how they freely chose.

    The brain can 'feel' stuff through the senses. I touch a hot or cold surface and I feel the heat or cold. But I cannot 'feel' my neurons working. Intracranial probing confirms that the brain cannot 'feel' itself. What does this mean for human brain activity? It means that whenever we have a conscious feeling that we have made a free willed decision, we cannot detect the brain activity that caused it, all the sub-conscious physical activity of the brain.

    Ask of a 'robot' advanced enough to be able to monitor its own software running, but unable to 'feel' the physical substrate upon which it was running, unable to feel the microcircuits at work. What would it feel like for such an automaton? Would it feel to it as if it had free will? Are we not biological automata that have self-awareness, self-monitoring capabilities that can read and feedback signals, such that we feel as if our decisions are arising free of any physical cause?

    Take a visual illusion, such as the rotating mask. No matter how we look at it we suffer the illusion that the concave side is convex. We know intellectually that the concave side is in fact concave, but we can't help 'seeing' the illusion, as if the mask is convex. But note, all the reflected light reaching the retina is showing us a concave mask - the light doesn't suddenly change to lay down a convex image on the retina. It is the human brain's propensity to see faces that makes the brain 'see' a convex face. Though we call this a visual illusion it is really a mental one. We have the advantage hereof being able to examine the mask from a changing point of view (POV) as it rotates in front of us.

    The problem with the 'free will' illusion is that it is an entirely internal mental illusion, with no data from the senses to help us with our intellectual assessment of it. And it is so well ingrained in humans, possibly an evolutionary convenience, and efficient mechanism, that even those of use that think free will is an illusion still suffer the illusion - just as we all suffer the mask illusion.

    "Those who subscribe to the reductionist school of thought would say that we do not have free will. According to this argument, a complete understanding of the world ..."

    You address here not just reductionism but determinism, which is a sort of classical physics notion that everything can be determined from its prior states. You miss the point of this argument. The point is that IF the universe were entirely deterministic then yes, everything would indeed be determined, pre-determined (not to be confused with the spooky 'pre-destination' or 'fate' which imply that there is some demon or god determining our outcomes - quite different and certainly not what a 'determinist' is arguing). Determinism rules out free will. So, then we ask, what is our universe really like? And we get quantum stuff added in, which then seems to make the universe un-deterministic, indeterminate. But that doesn't rescue free will at all.

    ...

    ReplyDelete
    Replies
    1. well done. put down everything that i wanted to say, but saved me the time, as i am late for work.

      most important bit: how in the nine hells are random number generators supposed to help rescue the concept/construct of classic contra causal free will?

      and: to really blow your mind, look up block time/eternalism. sort of makes the whole free will kerfluffle look pretty wee.

      Delete
  6. ...

    "there's one and only one corresponding state afterward, and that means only one choice"

    No. If there is only one possible state for some outcome then there is NO choice. But no matter, since this still misses the point.

    The significant point with regard to a determinate and indeterminate universe is that they would both be indistinguishable to us humans. Suppose the universe were truly deterministic, entirely, so that when it came into existence everything in it was in principle determinate to some outside entity. The significant question here is how would it seem to smaller entities that were part of that universe? Well, any component of a system is by definition a sub-set of the whole, both in its actual composition, and in its states over time. No component of a system (static or dynamic) can measure, record, monitor, perform calculations on, *determine* all the state of the system in which it exists. All the humans in the world with all the greatest computers they could muster would still only be a tiny fraction of the system they inhabit, and so could not possibly calculate all the states of the system sufficiently to any serious deterministic calculations. A system, a universe, is always indeterminate to components (e.g. humans) within it. So, your footnote is right. But this is also a limitation on those claiming to know they have free will!

    The consequence of all this is that if you claim to make a free willed decision you cannot possibly know that it is free willed, because you cannot do all the deterministic calculations that would falsify your claim - it is falsifiable. Simply saying, "I choose to raise my arm." and then doing so is not evidence of free will, because it could equally be evidence of prior physical causes.

    "unimaginative, crippled by the Einstellung effect" - The free willies are crippled by the introspective subjective feeling that our decisions are freely willed; and all modern philosophy that supports the free will position is heavily conditions by Descartes. Even among those that reject dualism, such as many 'compatibilists', they make all the noises of the dualist, by not offering any explanation of how free will might actually work.

    ...

    ReplyDelete
  7. ...

    "they are based on old paradigms and leave no room for new, "outside the box" ways of thinking."

    What new ways?

    The idea that 'free will is an illusion' is very specifically that we feel our will is free of physical causes, when it is not. "only the conscious aspect of free will ("I think I'll make a left turn here") is the illusory part?" But that's the point! That is the illusory part. You are confused toward the end of paragraph three. "because it assumes that only our conscious self is capable of making free decisions." Yes. That's what the free will claim is. "What if the actual free decision happens a half-second earlier in the subconscious" It does! But then it is still entirely physical, as determined by biological activity. So, to the confused result:

    "There's a lot more going on at deeper levels than it seems, and this is where free will may reside."

    How do you know this is where free will resides? By what mechanism? What is it that you are defining as free will now? What is the will? What is the will free of? Look at the term, 'free will'. It consists of the adjective 'free' qualifying the noun 'will'.

    We can make a stab at the 'will': The activity of the brain as an overall planning system that is driven by brain-body-external inputs in unbelievably complex system of untold number of feedback loops. The overall behavioural output of the brain-body system, that allows it to make short term decisions ("Mmmmm. Chocolate or vanilla? Vanilla I think."), or very long term plans, some of which work ("I want to be a brain scientist.") some of which don't because the odds are stacked against ("I want to be an astronaut.") , some of which don't because of conflicting internal brain-body decisions ("I must lose weight. But right after this chocolate bar.") But, that 'will' is still determined by the underlying biology, and by the chemistry, and by the physics.

    But what about the term 'free'? What is the above will, intention, brain-body mechanism, actually free off? The only sense in which the term 'free will' has ever made sense is in the Cartesian dualist sense, of the will being free of physical causes.

    The best we can say is that the 'will' is an emergent property of the brain-body system, and that in experiencing the brain-body in action, as we experience live those decision making mechanisms, the interactions of biological elements, chemical molecules, physical elements, the overall subjective feeling that this brain-body system has of itself is that it is making decision free of the physical causes that are actually there. The freedom of the will from its physical causes is an illusion.

    This is not as bold and unsupported a claim as is the one that we have free will. The inference that free will is an illusion comes from all we know from physics of never finding anything that is free of physical causes, and from evidence that shows what we think are freely will choices are often not, and from evidence that our subjective claims about ourselves are often fallible.

    ReplyDelete
    Replies
    1. True, free will is not rigorously defined. This makes it impossible to test whether an entity has it or not.

      I would define free will as complex, continuing response to stimuli, that is unpredictable to other entities equally complex (and possibly even entities that are far more complex). A sufficiently complexified Watson would pass a Turing test in that no human would be able to distinguish whether it is acting "freely," or not. Therefore it has free will according to that definition.

      Can we agree that, while not certain, it is likely that matter and energy interact fundamentally the same in living biological systems as in non-living, non-biological systems? Otherwise, there is something special about life in terms of physical laws, an idea few of us can get behind. This is how I arrive at the opinion that a sufficiently complexified Watson would have (what we call) "free will" -- even if the behavior doesn't satisfy what some would call truly "free," whatever that might be.

      Delete
  8. "The high-level executive functions of the brain, which include thoughts and sensations, are only a small part of consciousness, like the images displayed on a computer screen. There’s a lot more going on at deeper levels than it seems, and this is where free will may reside."

    The unimaginably complex workings of the brain are deniable. But how free are we a really when we are unaware or unconscious, as it were, to what's going on deep inside? Of course, I'm referring to free as in the ability to choose otherwise, in the same condition (in every condition that could be possibly imagined, that is.)

    Overall, does it not boil down to reactions? Everything has an initial reaction, and we don't "choose" the reaction that caused us to react.

    ReplyDelete
    Replies
    1. But I think we do -- we just have to give up on an old, classical definition of what it means to choose freely. Given the same stimuli, any living organism will not always react in the same exact manner, as a simple nonliving entity would. To me, that's all that's necessary to qualify as free will. Human free will is just an extremely complexified version -- so complex, in fact, that we lead ourselves to believe that free will is something that it perhaps is not.

      Delete