Is it possible to build a sapient machine ?

[Bugmaster]

Topic

Many threads regarding dualism, theism, and philosophy in general, often run into this topic. Is it even hypothetically possible to build a computer that will be sapient -- i.e., an artificial being that will think, feel, and socialize as humans do ? Well, here's our chance: to resolve this debate once and for all ! Smoke 'em if you got 'em, people, this post is gonna be a long one.

I claim that creating Strong AI (which is another name for a sapient computer) is possible. We may not achieve this today, or tomorrow, but it's going to happen sooner rather than later.

First, let me go over some of the arguments in favor of my position.

Pro: The Turing Test

Alan Turing, the father of modern computing (well, one of them), proposed this test in ages long past, when computers as we know them today did not yet exist. So, let me re-cast his argument in modern terms.

Turing's argument is a thought experiment, involving a test. There are three participants in the test: subject A, subject B, and the examiner E. A and B are chatting on AIM, or posting on this forum, or text-messaging each other on the phone, or engaging in some other form of textual communication. E is watching their conversations, but he doesn't get to talk. E knows that one of the subjects -- either A or B -- is a bona-fide human being, and the other one is a computer, but he doesn't know which one is which. E's job is to determine which of the subjects is a computer, based on their chat logs. Of course, in a real scientific setting, we'd have a large population of test subjects and examiners, not just three beings, but you get the idea.

Turing's claim is that if E cannot reliably determine which being -- A or B -- is human, then they both are. Let me say this again: if E can't tell which of the subjects is a computer, then they're both human, with all rights and privileges and obligations that humanity entails.

This seems like a pretty wild claim at first, but consider: how do you know that I, Bugmaster, am human ? And how do I know that you're human, as well ? All I know about you is the content of your posts; you could be a robot, or a fish, it doesn't really matter. As long as you act human, people will treat you as such (unless, of course, they're jerks who treat everyone like garbage, but that's another story). You might say, "well, you know I'm human because today's computers aren't advanced enough to post intelligently on the forums", but doesn't prove much, since our technology is advancing rapidly all the time (and we're talking about the future, anyway).

So, if you're going to deny one of Turing's subjects his humanity, then you should be prepared to deny this humanity to everyone, which would be absurd. Therefore, a computer that acts human, should be treated as such.

Pro: The Reverse Turing Test

I don't actually know the proper name for this argument, but it's sort of the opposite of the first one, hence the name.

Let's say that tomorrow, as you're crossing the street to get your morning coffee, you get hit by a bus. Your wounds are not too severe, but your pinky is shattered. Not to worry, though -- an experimental procedure is available, and your pinky is replaced with a robotic equivalent. It looks, feels, and acts just like your pinkie, but it's actually made of advanced polymers.

Are you any less human than you were before the treatment ?

Let's say that, after getting your pinkie replaced, you get hit by a bus again, and lose your arm... which gets replaced by a robo-arm. Are you human now ? What if you get hit by a bus again, and your left eye gets replaced by a robotic camera -- are you less human now ? What if you get a brain tumor, and part of your brain gets replaced ? And what if your tumor is inoperable, and the doctors (the doctors of the future, of course) are forced to replace your entire brain, as well as the rest of your organs ? Are you human ? If so, then how are you different from an artificial being that was built out of the same robotic components that your entire body now consists of ?

Note that this isn't just idle speculation. People today already have pacemakers, glasses, prosthetic limbs, and yes, even chips implanted in their brains to prevent epileptic seizures (and soon, hopefully, Alzheimers). Should we treat these people as less human than their all-natural peers ? I personally don't think so.

Ok, I know that many of you are itching to point out the flaws in these arguments, so let me go over some common objections.

(to be continued below)

[QED]

There's a difference, though. We can simulate ferromagnetism in a computer, however inside the computer the phenomena is a set of flags, that's in part why it is a simulation of the behavior of ferromagnetism and not really ferromagnetism.

Granted there are such things as simulations, but we have to be wary of confusing the issue by dividing the world up into arbitrary categories. It looks to me as our difference in opinion here might revolve around such a category error. Reductionism is a dirty word to some, but is eminently applicable in the case of biology. Now that we have the power to see material on the scale of individual atoms there is no longer an essential difference between the organic and inorganic. Organizational Information becomes the only real distinction and, just as information can be carried from one media to another, it would seem to me that any biological function should in principle be equally transferable.

Similarly, no doubt CPU chips can simulate robots having behavior as though they are in pain, but they are not really in pain. Give them logical access to those registers, and they can reset those flags such that their behavior is no longer displaying the actions of those subroutines.

Well, by now you must understand that I don't think you're justified in referring to such simple models -- they only present a caricature of what I am tyring to put forward here. It might be convenient for your argument that our trivial simulations fail to deliver any discernible amounts of qualia, but this does not provide sufficient grounds to reject the notion that sufficiently thorough simulations would render the same experience of pain. The fact that flags can be voluntarily reset merely represents an arbitrary modification to the simulation -- one that is functionally equivalent to the self-administration of pain killing drugs.

Notice, this is not the case for ferromagnetism. If we give a AI robot software control to its CPU and all registers (i.e., without mechanical access where they can disrupt the atoms of the ferromagnetic material), then the robot cannot stop ferromagnetism inside their mechanical bodies just using software control.

I'm not sure I get your point here. It sounds like exactly what I would expect given that you are preventing any interference with the source of ferromagnetism.

This is a reasonable analogy to pain inside our heads. The processes that create pain are physical processes (e.g., firing of neurons, chemical reactions, etc.), and instead of a ferromagnetic phenomenon emerging inside our brains, we have a pain phenomenon. We cannot shutoff this pain for the simple fact that it is a physical phenomenon (just like consciousness is a physical phenomenon). However, what separates ferromagnetism from pain is that ferromagnetism is brought on by the emergence of electromagnetic waves working in a collective action, whereas pain is probably brought on by the emergence of patterns within neuron firings, chemical processes, etc., working in a collective action. Both are physical phenomena that can be mapped into a geometric system, but pain requires a different conception of what physical is. We can use the concept of a "virtual process" as an anagolous concept, however in this case it applies to the real world.

Again, I don't see any contradiction to the view that pain is a physical phenomenon that will emerge from any similar collective construct.

As I alluded to, I think you're confusing micro-cognition as a physical phenomena (which I agree with), and micro-cognition as a computational property of AI. Simulating a brain inside a computer probably would not give you a conscious individual that can experience pain. However, building a brain that can produce brain waves, etc., might succeed if we had the resources to do it along with knowing enough about how such processes occur in nature.

But this seems to call on as yet unknown mechanisms. I agree that this might turn out to be the case, but what I am pressing for here is a point that could easily be totally overlooked and that is we might need look no further than the collective effect of particles and fields just going about their ordered business.

Modern technology has forced us to reconsider other age-old distinctions such as those between man-made (synthetic) and natural materials for example and I'm always on guard for people being misled as a consequence.

[harvey1]

Well, by now you must understand that I don't think you're justified in referring to such simple models -- they only present a caricature of what I am tyring to put forward here. It might be convenient for your argument that our trivial simulations fail to deliver any discernible amounts of qualia, but this does not provide sufficient grounds to reject the notion that sufficiently thorough simulations would render the same experience of pain.

In my view, what provides us sufficient grounds to reject a computational view of the mind is that in principle it cannot be shown how subjective experiences (e.g., qualia, consciousness, intent) arise from these computational processes. These are based on syntactical manipulation of language that fail to show how semantics is feasible.

is functionally equivalent to the self-administration of pain killing drugs.

Bugmaster raised a similar objection, however I don't see how you can show that as the case. Drugs appear to dilute the pain but they do not shut off any flags. I imagine that there could be chemicals that could retard ferromagnetism, but would not succeed at shutting it off completely.

I'm not sure I get your point here. It sounds like exactly what I would expect given that you are preventing any interference with the source of ferromagnetism.

My point is that consciousness is a physical process, and not a computational process.

Again, I don't see any contradiction to the view that pain is a physical phenomenon that will emerge from any similar collective construct.... But this seems to call on as yet unknown mechanisms. I agree that this might turn out to be the case.

I don't think it does point to an unknown mechanism since we already have plenty of examples of how self-organizational processes bring about a kind of "consciousness." For example, if you observe how birds flocking together causes the flock to act and behave as one ogranism, then you would observe emergent "conscious" behavior.

[QED]

Well, a very interesting thing happened to me today: I will try and spare the sordid detail but each day (for about a week now) I have to show up at my doctors surgery and (with no anaesthetic!) have something poked into my back Now the nurse that does this is very nice indeed, but it feels damn awful and takes around 15 seconds. In this time I've been pondering this whole question about pain and today I noticed that I can actually think it away -- well not so much away as turn it into something else. The process started by thinking that instead of pain it was just a peculiar feeling. Then I got from it being peculiar to it being like a strange screeching sound. I know this sounds pretty wacky, but I'm almost beginning to look forward to experimenting the next time I go (I'm being given the weekend off for good behaviour).

Now with this and another very recent experience in mind (I was trying to solder a resistor in series with a PC cooling fan and while not having enough hands to hold the wires, solder and iron, half expecting to burn myself, my left hand brushed against the cold metal case and I immediately thought I had burnt myself) I can see that the experience of pain is not as absolute as your ferromagnetic field. Indeed it seems to me to be an interpretation of certain nerve states and this interpretation is subject to expectation. Anyway, if nothing else this little debate has actually helped me face the music at the quacks - and for that I am really quite grateful

[harvey1]

I noticed that I can actually think it away -- well not so much away as turn it into something else. The process started by thinking that instead of pain it was just a peculiar feeling. Then I got from it being peculiar to it being like a strange screeching sound. I know this sounds pretty wacky, but I'm almost beginning to look forward to experimenting the next time I go (I'm being given the weekend off for good behaviour).

Ouch... Just hearing you talk about pain has me squinching...

Although I don't want to deter you from probing the depth of the feeling inside your mind, I think it is fair to say that you are encountering something real and tangible. I haven't experienced a great deal of pain recently, but when I have, I have probed the same kind of questions. Where is it coming from, is it real, is there a flag that I can turn off, etc.. My conclusion is that it's as real to me as ferromagnetism.

[Bugmaster]

Ouch QED, as much as I love science, I hope that you get better soon, so that your experiments with pain could cease :-/

Harvey,

I haven't experienced a great deal of pain recently, but when I have, I have probed the same kind of questions. Where is it coming from, is it real, is there a flag that I can turn off, etc.. My conclusion is that it's as real to me as ferromagnetism.

Is it as real "to you" as ferromagnetism ? Or is it as real as ferromagnetism, period ? There's a big difference there. For example, the Earth is quite flat "to me" most of the time, but, in reality, the Earth is round. Again, you're venturing into solipsism country here.

Furthermore, QED has demonstrated that he can turn pain on and off like a flag -- ok, he can't turn it off, but he can redirect it. I'm sure he's not alone in this.

[QED]

Well thanks for your sympathy fellas, I sort of regret writing about it now but I'd just got back from the quack and settled down in front of the computer with a nice glass of bourbon and, well, it all just spilled out (not the whiskey! ).

If I think in terms of pain as being either real or virtual, I'd have to say that it's virtual. And there are so many different flavours of feeling (why did we have to zero in on the most unpleasant?) that it simply doesn't have the ring of being a fixed phenomenon like ferromagnetism. It seems to me to have more to do with some sort of (dis)harmony - a periodic wave that is handled in a similar way to our auditory processing. After all, music is perceived as series of emotions. Musical appreciation seems to me to be about recognition (matching what we hear with things in our sound libraries) and associated emotions are released as a result. I also think (pet theory warning!) that the library might be partly stocked when we are born, but that's material for another debate!

[harvey1]

So it sounds like you're asserting that consciousness is a separate entity, perhaps even a law of nature -- sort of like gravity or ferromagnetism.

What I'm asserting is that consciousness shares some features of the physical world as ferromagnetism (and probably gravity too). Namely, it is an emergent phenomena based on a lattice that at some critical point reorganizes into self-organized structure having new properties than prior to the critical point. There is a great deal of evidence for self-organizational principles, and I'm shocked that you don't see the relevance of these principles with regard to human consciousness.

If that were true, then your ferromagnetism analogy would make sense. However, QED and I explicitly deny your assertion (ok, at least I do, I can't speak for QED); I don't believe that consciousness is an independent entity (and I don't believe that laws of nature are Platonic entities, either, but that's another story). Instead, I believe that consciousness is just computational activity (be it brain-based or computer-based) that leads to a certain set of behaviors. So, your ferromagnetism analogy does not apply to my argument.

I know this is what you believe, but the point of this thread is to show that this view will naturally lead to p-conscious machines. I thrown down a few challenges here (e.g., programming pain), and you haven't shown how in principle this is possible without relying on some magical antedont.

Since you're proposing that consciousness is more than just computational activity that leads to certain behaviors, the burden of proof is on you.

How so? I am merely giving my opinion as to what human consciousness really is, and therefore it helps to explain why I think computational strategies are completely off-base. Of course, if you want to show how pain is programmable in principle, then we don't have to listen to my little speculations...

I've challenged you repeatedly to show me that you can detect my consciousness without relying on observing my behavior; so far, you haven't really come with a good answer. If the existence or absence of consciousness makes no measurable difference, then it becomes an empty concept, and we should focus on behavior in order to determine the subject's humanity.

I'm sure you realize that methodological and logical behaviorism have been disavowed by just about everyone having anything to do with the study of the mind, right? So, why are you pursing a belief that is just not accepted nowadays?

[QED]

Namely, it [consciousness] is an emergent phenomena based on a lattice that at some critical point reorganizes into self-organized structure having new properties than prior to the critical point. There is a great deal of evidence for self-organizational principles, and I'm shocked that you don't see the relevance of these principles with regard to human consciousness.

Sure, but then what's the significance of emergent phenomena anyway? That is the point I'm trying to make here. If we take any other emergent phenomena I think it's fair to say that, in principle, it's always understandable (in computational terms) from the interaction of its individual components. Of course this nearly always gets obscured by the practical limits preventing us from performing such computations. This might lead some people to think that magic is taking place, but in reality it's just that we aren't smart enough to keep track of everything that's going on.

In this way I think you're artificially breaking the link between consciousness and the components from which it arises -- when in fact the link is intact all along but obscured by our inability to perform the computations required to follow it. This enables you to challenge us to come up with some "missing link" which does not actually exist. Hence when you say:

Of course, if you want to show how pain is programmable in principle, then we don't have to listen to my little speculations...

It's simply not an appropriate challenge.

[harvey1]

Sure, but then what's the significance of emergent phenomena anyway? That is the point I'm trying to make here. If we take any other emergent phenomena I think it's fair to say that, in principle, it's always understandable (in computational terms) from the interaction of its individual components. Of course this nearly always gets obscured by the practical limits preventing us from performing such computations. This might lead some people to think that magic is taking place, but in reality it's just that we aren't smart enough to keep track of everything that's going on.

I think this flies in the face of dynamical systems where we know that the emergence being discussed is more than epistemological emergence. For example, this paper
briefly summarizes the immense impact that Ken Wilson's formalization of renormalization of dynamical systems has had in physics. This doesn't even scratch the surface for ecology, economics, evolution, etc., along with the universality that all these dynamic systems share. Now, surely you aren't saying that symmetries aren't really broken with respect to the important gauge symmetries, right? I mean, don't you agree that photons are ontologically distinct from W and Z particles (even though they are the same particle prior to the spontaneous symmetry breaking that occurred at the end of the electroweak epoch)? If you agree, then how could you possibly say the particles are differentiated for reasons other than spontaneous symmetry breaking that acts as the cut-off between being an electroweak unified particle and being a photon, W, or Z particle?

My problem with both your's and Bugmaster's approach is that it doesn't seem to recognize the importance of dynamical systems studies, and nor does your view account for the universality of how it is that all of these systems (physical, chemical, biological, social, etc.) can be explained using the concept of symmetry breaking (both continuous symmetries and discrete symmetries).

What dynamical systems studies reveal to me is that we are justified in claiming that emergent systems that undergo phase transitions at critical points (via symmetry breaking events) are novel systems. If the brain and p-consciousness are dynamical systems, then given the universal nature of phenomena studied in dynamical systems, we have no reason to doubt this conjecture, then why are we not equally justified in claiming that consciousness is an ontologically emergent phenomenon?

[harvey1]

How do you know I feel pain ? Can you see the qualia inside my astral head ? Or is it through my behavior (such as screaming "ouch", for example) ? If so, then how will you distinguish me from an AI that acts exactly like me ?

This is irrelevant. You just said that you agree that you feel real pain. Now I want you to explain how this function is possible if you really feel it. In actuality, I don't have to justify to myself that you feel pain since I feel pain, so I already feel justified in needing an explanation for pain.

Hmm, that sounds a lot like solipsism to me. You don't feel justified in believing that anyone has any pain at all, except for yourself. If you're willing to go that far, we can discuss it further, but I personally think that solipsism is absurd.

I see that response as a Red Herring. The subject matter for my question about your pain has nothing to do with whether I am justified in believing that you have pain or not. (I think that I am justified in believing that others, including QED and yourself, have pain so therefore I see no reason to explore whether this amounts to a solipsist belief on my part.) The subject matter for my question is whether you believe you have pain. If you agree that you do (which you did), then you should explain that experience in terms of an algorithm that causes that feeling. Just referring to algorithms that imitate pain reactions is not explaining your pain. Don't worry why I believe you. I have my reasons for believing that you really experience pain. It has nothing to do with some kind of solipsist tangent. So, if you don't mind, reply to this question: can you describe in principle how the algorithm running inside your head gives you a feeling of pain? If not, then why are you so sure that future AI programmers can program it with standard computer logic? It seems like faith on your part. Since we share different faiths, I need reasons that we both can share. How do you convince someone that pain is nothing more than a computation who doesn't already share your faith that pain is a computational algorithm?

If "pain" is a label we give to marble bouncing events, and you've identified the exact marble configuration that leads to pain, then you've clearly shown how pain really occurs. I don't see a problem. Yes, actual pain that occurs in real live humans is a lot more complex than a handful of marbles, and so I can't "explain why it is that I feel pain when I am hurting". However, a modern neurobilogist would be able to do this... You should ask one of them.

I haven't heard of a good "in principle" explanation on why it is that we experience pain. So, that's why I'm asking you. If you don't have one, then why do you think that this is something that adding enough complex programs will somehow make it appear?

This statement strikes me as ludicrous for a few reasons. One, I don't have to justify ferromagnetism since it is an observed property in nature.

Er, and consciousness is not ? Be careful how you answer. If you say that consciousness is not an observed property in nature, then you make it irrelevant. If you answer that it is indeed an observed property in nature, then you have to tell me how I can go ahead and observe it, and you haven't been able to do that so far.

What do you mean by observe? In my view, the term "observation" is based on what humans accept as a matter of convention. So, for example, if we sent a lone astronomer to Mars, and they witnessed a meteor hit the horizon, then this is an observation. If the same astronaut observes that they have a conscious state, then this is also an observation. Since the far, far majority of us all have a very similar experience of what it means to be p-conscious, we can consider our experience as an observation. In fact, all scientific observations are based on the fact that we have reliable observable capabilities--otherwise we could never agree even on the most basic aspects of scientific investigation.

I actually count strings as being physical as well... why wouldn't they be ?

Strings are theoretical entities and beyond observable status.

If the algorithm for pain is, "if(nerveSignal > threshold) { ouch(); }", then the opening of an extra logic gate -- the one that controls the high bit of nerveSignal, for example -- can indeed be directly traced to pain.

So, "ouch()" is a flag. But, again, how is it that you have the feeling of pain versus the logical indication of pain? I need to see the code for the feeling versus the flashing light indicator.

however the computer didn't have to have this feeling programmed into its instruction set to behave just like you.

Why not ? If the computer is a copy of me, then it has the same
things "programmed in" that I do, right ?

If I recall correctly, the computer program is just mimicing you. Why do you think an outward mimicing of you entails that it share your inward feelings? Are you saying that because it mimics you that it must share your inward feelings? Why should this be the case? Can't you have a puppet portraying a boy (controlled by strings) that mimics human actions without the puppet actually having a boy's emotions and feelings?

I think that a "mental process" is just an abstraction we use to explain human behavior. Thus, if we think someone acts human, then we are justified in treating him/her/it as such. When looking at biological humans, we can actually trace their behavior to physical changes in their brains; when looking at electronic humans, we can trace their behavior to physical changes in their circuits. So, yes, their mental processes are identical to their physical processes, but those are all just abstractions we use to describe them.

BM, do you realize that this is behaviorism? It has been fully invalidated. I would say that there are no professional behaviorists left. At first I thought you were an epiphenomenalist since I misinterpreted a few of your earlier statements.