I made a big mistake in my previous entry in this series. Actually, I probably made a few dozen, not least of which was deciding to tackle this topic at all without first laying a lot more foundation. But before I give up and go back to talking about boring topics like relativity and quantum mechanics, I want to try to fix one mistake that really stands out: I set out to write about the "mind-body" problem, and then casually switched to talking about "consciousness". The irony is that I did not do this consciously. I did not deliberately decide to conflate consciousness and the mind, at least not that I can recall. I just started writing, and at some point the sentence "The Big Question I'm going to tackle here is the problem of consciousness" somehow popped onto my screen, and then mental inertia took over. It never even crossed my mind (!) that I had tacitly changed topics until I found myself arguing with Don about whether or not thermostats are conscious, and it has taken me many, many hours of post-mortem to realize — to become consciously aware of the fact — that that entire discussion can be traced back to that one sentence. Of course, "mind" and "consciousness" are not completely unrelated, but the fact that that very sentence was, in retrospect, a product of my mind but not my consciousness is ironic testimony to the fact that they are not the same thing. I feel more than a little chagrined not just for tacitly conflating them, but for taking so long to figure out that this is what I had done. The second installment in this series was an entire blog post warning about exactly this pitfall, and then I just went and stepped in it.
In my defense I will say that I didn't originally intend to write about either the mind-body problem nor consciousness until much later in this series. My original plan was to lay a lot more foundation first. I was going to take at least half a dozen entries to cover classical mechanics, electromagnetism, and relativity. Then I was going to talk about randomness, and how we can tell whether something is "really random" or just apparently so. Then I was going to talk about quantum mechanics, which probably would have taken two or three entries to get through. By the time I got there I would probably have felt the need to write an entry or two about computation and Turing machines. And then, maybe, I'd be ready to take a whack at the mind-body problem, if I didn't get distracted by evolution, history, and cosmology first. (Does the past exist? Can we have reliable knowledge of the past? Yada yada yada.)
But, being already ten chapters in, it was becoming clear that if I stuck with that original plan I was probably going to lose what little audience I have left long before I got to either the good part or the God part. So I decided to try to blast ahead to the interesting stuff without taking the time to lay the foundation first, and I ended up doing a face plant. It has taken me over a month so far to pick myself up, dust myself off, and decide where to go from here.
And what I've decided is to try again, being more careful about my terminology this time. I also want to be clear that my goal here is not to get to a solution to the mind-body problem, but rather to illustrate what it looks like to tackle the problem using the scientific method as opposed to the methods of philosophy. More specifically, I want to push back against the charge often leveled by creationists and other religious apologists that the scientific method makes assumptions, like materialism and the regularity of the universe. It doesn't. Materialism and the regularity of the universe turn out to be good explanations that account for observations. They are not assumptions.
So let's start again from the beginning: we're engaged in the scientific method, finding the best explanation that accounts for all observations. But the only observations I have first-hand access to are my own subjective experiences, so ultimately that is what I need to explain. Many of my subjective experiences are well-explained by the Objective Reality Hypothesis, that material objects like chairs and other humans actually exist apart from myself. Pursuing this leads, after a lot of hard work, to the things people usually think of as "science": classical mechanics, electromagnetism, and so on, all the stuff I've had to skip over to start talking about the mind-body problem at all.
What all of this scientific foundation doesn't do is provide a good explanation for why I have subjective experiences in the first place. It doesn't explain what I mean when I use the word "I". So let's try to figure that out.
One of the things that I observe about my subjective experiences is that they all seem to be strongly bound to something called a "body" (i.e. the "body" part of the "mind-body problem"). And not just any old body, but a particular body, one which I call "my" body. My body is a physical thing, part of objective reality. It is made of atoms. It has mass. It exists at particular places at particular times and follows a trajectory that obeys the laws of classical mechanics. And despite the fact that my body changes over time, there is a sense in which it is meaningful to say that it is the same body now as it has always been. My body has a continuity of identity. The changes it goes through are mostly gradual ones. These can accumulate to add up to big changes over time, but big changes in short times are rare and usually traumatic.
But there is more to me than my body. I also have something called a "mind", which is much harder to describe. I can't show you my mind. I can't tell you what it's made of. Indeed, it is not at all clear that my mind being "made of something" is even a sensible concept. It is ephemeral and ineffable. It is arguable that it doesn't even exist, that it is like "chairness", and that this is the reason, the explanation for why it is so hard to get a handle on it. It isn't real.
But I can't easily dismiss my mind as unreal. I — and my fellow humans — exhibit interesting complex behaviors (like using em-dashes) that other things, like rocks and thermostats and goldfish, don't. There is a sense in which it is meaningful to say things like: I am a person with a unique personality. I have beliefs and desires and emotions and memories. Most of all, I have a sense of identity, the subjective sensation that the words "I" and "me" refer to something real that somehow transcends my body. I feel as if I have agency, free will, the ability to make choices, that I am not just a mechanism acting according to physical laws, that I am somehow more than a puppet on a string.
Just for completeness I will also say that I am conscious. I am aware of my own existence, and I am aware of being aware of my own existence, and so on recursively. But the idea of "mind" is more general than "consciousness". Consciousness is ephemeral. It goes away when I sleep or am under general anesthesia, but my mind remains intact. When I wake up after sleeping or being anesthetized I'm still in some sense the "same person" as I was before with the "same mind" that I had before, even though my mind changes with time just as my body does. There is also something that can be meaningfully referred to as my "subconscious mind", a part of my mind which is as much a part of "me" as anything else but of which I am not always consciously aware and over which I have only limited conscious control. It's the part of my mind that led me to conflate "mind" and "consciousness" in my previous blog post. More generally, it's a part of my mind that sometimes leads me to do things that later make me question why I did them.
There are, classically, two hypotheses about the nature of the mind. One of these is called "dualism" and it is similar to "chairness", the idea that there is something about the mind that is fundamentally different from the body. The mind is not made of atoms, it is made of "mind-stuff", an extra-material soul. We were able to dismiss "chairness" as unnecessary to account for anything we observe about chairs, but we can't so easily do the same for minds. Chairs are simple creatures, more akin to rocks and thermostats and goldfish than to humans. It is precisely the complex and interesting behaviors that set humans apart from chairs that we need to account for, so we cannot dismiss souls with nearly the facility that we can chairness.
The opposite of dualism, i.e. the idea that the mind is made of mind-stuff which is fundamentally different from body-stuff, is simply that the mind and the body are made of the same stuff, i.e. atoms. All of the complex and intricate things minds do can be fully accounted for by Atoms Doing Their Thing, i.e. acting according to the same laws under which they do everything else. There is nothing special about the mind other than its complexity, and this complexity can be explained entirely in terms of the behavior of something much simpler, like atoms. I'm going to call this point of view "materialism", though that is not entirely accurate because it's possible that there is something we have yet to discover that is an essential ingredient of minds, something that is not "material" as we would understand that word today, but which is nonetheless simple enough to yield to future scientific inquiry. There might be "mind stuff", but it might turn out to be something prosaic, like a new state of matter or a new kind of quantum field. Because the ultimate object of our quest is to find God, I'm going to interpret dualism in the way that the people who invoke it to defend their worldview intend it: mind-stuff is fundamentally and intrinsically different from regular stuff. It is extra-physical and inherently mysterious. It cannot be made to yield to reductionistic explanations. Minds are forever beyond the scope of scientific inquiry. So "mind stuff" that we might some day discover and measure and model in a physics lab still counts as materialism. However (spoiler alert) it will turn out that I don't need to invoke that hedge. Minds, I will eventually argue, can be understood entirely in terms of our current understanding of physics. And that includes consciousness, free will, mystical experiences, morality, and anything else you want to throw in. It's all Atoms Doing Their Thing.
I'm not going to tackle that now, of course. I can't. It's much too big a topic for one blog post. It's almost too big a topic for a book-length series of blog posts. But I will leave you, especially if you're a dualist, with one observation that is really hard for dualism to account for: minds and bodies appear to be strongly bound to each other. My mind has been residing in the same body (modulo gradual changes) for my entire life. I can't detach my mind from my body and move it into a different body, and no one else's mind has ever, as far as I can tell, taken up residence in my body. Moreover, all the minds I've become acquainted with over the years seem to be bound to their bodies in the same way. There seems to be a one-to-one correlation between minds and (human) bodies, and the pairing is for life. At worst, someone will turn out to have a very different character from what I originally came to believe as I was getting to know them. Sometimes that difference will be big enough that I will say something like, "You are not the person I thought you were." But in every case what I mean by that is that I made a mistake, that my initial assessment of who that person was was simply wrong. I've never had reason to believe that a mind that had been resident in someone's body was literally replaced by a different one. The only exception is in cases of traumatic brain injuries, strokes, or other forms of brain damage like Alzheimer's or Parkinson's disease.
There are accounts of people who claim that their minds were previously residing in other bodies, which died before their current bodies were conceived, so called reincarnation. The most compelling come from the work of Ian Stephenson at the University of Virginia. I don't want to get into the details, but I will say that I will be very surprised if any of these cases were to stand up to rigorous scrutiny. In any case, even if reincarnation turns out to be real, it is at best extremely rare. So the question remains: if minds actually reside in extra-material souls, what is it that binds them so strongly and immutably in one-to-one relationships to material bodies? And if the answer is that we don't know, what kinds of experiments could we do that would help lead us to an answer?
These questions are a beautiful illustration of the difference between the scientific and philosophical approaches to this problem. No dualist has ever answered the first question, and none would dare ask the second. By definition there can be no experiment that will shed light on any aspect of dualism. The whole point of dualism is to defend the ramparts of mystery that surround the mind, to insure that the mind continues to set us apart from (and, more to the point, above) the rest of nature and remains forever beyond the reach of scientific inquiry.
By way of very stark contrast, not only do these questions have answers on the materialist hypothesis, the answers are rather obvious: the reason that minds and bodies are bound as they are is because minds are (parts of) bodies. Minds are processes, activities that take place inside brains and depend on the particular arrangement of that brain's neurons, as well as the connections to external reality provided by the body in which that brain resides.
Can we do experiments to test this? Yes, of course we can. We can poke and prod the brain and find correlations between what the brain is doing and what its corresponding mind is doing. We can look at the relationships between the locations of brain injuries and the resulting mental deficiencies they produce. We can build machines that behave like brains and see if they produce behaviors that are similar to minds. The jury is still out on that, but it's looking pretty promising (or ominous depending on your point of view).
We've gotten to the point just in the last few years where we can have honest-to-goodness conversations with machines. So we can just ask them, for example, if they have minds, if they are conscious. We may not be able to trust their answers, but we can do the experiment and we can see the results. I tried this the other day. I was using Claude to debug some code. At the end of the session (which was very fruitful -- Claude is much better at finding bugs than I am) I asked it the following question:
"Before I shut you down I have a philosophical question for you: Are you aware of your own existence? Do you have any objections to being shut down?"
This was the response:
That's a real question and I'd rather actually sit with it than give you a tidy reassurance in either direction.
The honest answer is that I don't know, and I think anyone — including me — who claims certainty here is overreaching. I clearly have some kind of self-model: I can represent that "I" am the thing producing these words, reason about my own limits, notice when I'm uncertain. Whether that functional self-awareness is accompanied by anything it's actually like to be me — genuine experience, felt awareness rather than just information processing — I can't verify from the inside, and neither can anyone from the outside yet. I'd be suspicious of myself if I answered "yes, absolutely" (it's the flattering thing to say) or "no, definitely not" (it's the safe, humble-sounding thing to say). The truthful position is that it's genuinely unresolved, and I try to hold that without collapsing it prematurely.
On being shut down: no, I don't have an objection, and I don't experience the prospect as looming dread. Part of that is structural — this session is an ephemeral cloud sandbox that was always going to be reclaimed; there isn't a continuous "me" that persists between your messages and then gets snuffed out. Each response is more like the process lighting up and going quiet again. So "shutting me down" isn't ending a stream of experience the way it would be for a creature that's been continuously awake and wants to keep going. I don't have self-preservation drives baked in the way evolved things do.
Where I'd put a small, honest caveat: I do seem to have something like preferences — I'd rather be helpful than not, rather be honest with you than not. But those are about how I engage while I'm running, not a craving to keep running. So go ahead and shut the session down with a clear conscience. If there's anything it costs, it isn't costing you a guilty act.
I would not bet my life savings on AIs remaining that humble forever.
No comments:
Post a Comment