Our class discussion the other day highlighted Searle’s main point, that while computers may pass the Turing Test, it is still not a valid test of consciousness.  Searle’s argument leaves me wondering about the features that would be necessary for a computer to have limitless creative capacity.  When I think of our Scribbler lab I remember that the closest my Scribbler could get to creativity was randomness followed by a programmable action—which in my opinion is still very far from real and conscious creativity.  Placing creativity on par with consciousness, I find Searle’s argument most compelling in that the Turing Test cannot determine the consciousness of an entity, despite the fact that it may be possible for a computer to pass the test.

It is difficult to side with either Brooks or Searle about whether or not it will ever be possible to duplicate the mechanical properties the brain as well as the rules that govern its actions.  Unlike Searle, I’d like to think that AI is possible at some level yet I will take from him that the Turing Test is not the proper benchmark of determining whether or not a machine is intelligent.

Imagine a woman named Mary is in a room and she has had an operation where she can only see in black and white. She has access to all the cutting edge medical technologies and research and becomes the world’s foremost expert on color vision. In particular, she knows everything there is to know about seeing the color red (wavelengths, rods, cones, etc). No matter how much she knows, however, we have an intuition at least that if the procedure is reversed and she is taken outside the room and sees something red, then she will learn something new namely what it MEANS to see the color red or what red looks like.

Some argue that Mary will not learn anything new when she first sees red in which case artificial consciousness seems plausible. However, if she does learn something new then there seems to be something about qualia intimately bound up in consciousness which does not bode well for AI. I am not sure if this connection I am drawing is exactly correct, but it is what I found myself thinking when I was reading Searle’s article.

One of Professor Arora’s lecture slides also brought up an interesting point: surely we would all concede that human consciousness is a strong form of consciousness. There are other life forms that we would admit to be conscious, but in a much weaker sense. Could then computers become conscious in some weaker sense? I think this has to seem much more plausible than a computer passing the Turing Test. However, the difficulty becomes how would we test if a computer has become conscious in this weaker sense? Could the Turing Test be extended to say that if humans and some examples of whatever given animal cannot tell the difference between that animal and a computer, then a computer has attained that level of consciousness? On an intuitive level this seems wrong (it is no stretch to imagine a robot that might fool a dog into thinking it is a dog, and because humans are not well acquainted with the consciousness of dogs, also fool a human as well), yet it is hard to think of a better analogous test to the Turing Test.

Anyways, sorry to ramble, just thought I would share some of those reactions to Searle’s article and our larger AI discussion which I have really enjoyed.

To me, there is a very real difference between the abilities of AI to convince people that it is “human” and actually proving its ability to think. To be human is very much contingent on thinking and reasoning, two factors that make up general speech and conversation. Sure, you could probably hold a respectable conversation with a strong AI machine, and maybe that machine could even pass the Turing Test…but what does this mean for society at large? In my opinion, it does not mean much. I think that the Turing Test is an oversimplified and convenient way of seeing if AI is “smart”. As I sat in lecture the other day, I couldn’t help but think about the number of incredibly smart people in the world who would not pass a Turing Test. Does this not make them real or human or intelligent? If anything, it makes them incredibly more human than the oversimplified machines that exhibit AI. It is a confusing and messy thing to try to determine, but I will certainly maintain that a machine will never reach the level of human interaction and intelligence that occurs in real beings.

Now, I would like to address a few issues we talked about in class. The argument that the Turing test is a fair (no obvious way to cheat ) one is not really convincing. Although there are a lot of different combinations of topics, it is relatively simple to produce relevant databases which comprise of the variety of topics a “normal” person would talk about, and thus simulate a Turing test passing computer. It is in fact even possible to create a program which learns from the people it talks about and imitates ways of writing and stylistic patterns. There is, however, a “slight” difference in the word “thinking” - in my opinion, it should relate more to the class of feelings than to the class of actions.

In conclusion, although the matter is somewhat philosophical, it is, at the current level of science, impossible to program a computer to “think”. This is, however, not at all discouraging as computers can be programmed to do almost any task (including decision making, using algorithms or approximations) any human can do and, further, tasks which will require immense physical or intellectual work (computations and a lot more). Thus, we don’t really need the computers to be able to “think”.

First, are humans machines as Brooks suggests? All of us in CS116 think we are not. One property of a machine is that it is manufactured —you can buy as many identical copies as you can afford. But as we discussed in class, if human cloning ever happens, this picture starts getting really muddled. That is one of the reasons most of us find cloning abhorrent. On the other hand, people have cloned sheep and dogs already, so how far off is human cloning?
Second, I want to raise the problem of the hypothetical bunny rabbits once again. What if there had been two intelligent species on earth? Dolphin lovers, or chimpanzee researchers will argue that there already are. But assume this other specie is so advanced it actually can carry on a conversation with us. We would obviously not need to consider them human or want to date them (at least I can’t imagine dating bunny rabbits). But how would we react to them? Speaking for myself, I can’t be sure.
I fully grant you that the bunny rabbit situation may be too hypothetical. We may be the one and only example of intelligence in the universe for all time.

To sum up, I hope it was useful for you to confront all these possibilities for a few hours or days. Unlike Brooks or Searle, I am not vested or fixed in my views on AI. If I start thinking about it, things get very confusing very fast. (Cardero, Mike, Lily etc., you are not alone.)

When we were asked to identify the main argument that Searle makes against AI I initially felt that his article really highlighted the difference between syntax and semantics. Perhaps that is what struck me as the most interesting point he made since for me, that really accentuates the difference between us humans who have the capacity for both while AI only syntax. After our class discussion, however, I realized that while that may be an issue, it’s not the real problem. The Turing test is the problem. Since the Turing test is used to try to determine the difference between human and AI behavior, it is in fact very possible that a computer could do something as well as a human and that an expert would not be able differentiate between the two. The real difference between human behavior and AI is the principle of thinking and the actual thought process that goes on in a human mind. This is just never going to be the same in AI, and the Turing test is thus not a valid test. Brooks approaches the topic from a very different perspective, which in my opinion is too simple. You cannot just look at a human as a body, you must realize how complex the mind is and realize that no form of AI will be able to replicate the intricacies of human thought process.

Here I am inclined to disagree. I would not deem a computer conscious even if it could respond in this way. Quite to the contrary, I would still resist categorizing his device as a conscious being. Prof. Arora’s response is “how do you tell this apart from your friends? How do you judge them except by input/output responses?” Here is where the reasining becomes trivial in my mind–we judge everything rational through input/output reponses. There is no part of the world that we do  not analyze by what causes we create (or see created) and the effects they produce. But therer is more to the human experience of the world than rigidly rational cause/effect, input/output analysis. The irrational is in many ways our chief means of engaging with the world, much as we would like to think otherwise.

Professor Arora asserts that he could find a machine to be his friend, but could not marry it (or as I interpret, could not love it). But that is really the judge of consciousness to me…not a rational, text-based decision of “do I think this machine can process me felicitously”. Am I capable of loving this entity outside myself? Do I think that it could potentially satisfy me emotionally as well as intellectually? That is my Turing test. That is what would allow me to judge a computer conscious.

I find Brooks’ argument towards the end of his book fascinating because he speculates on the reasons why we are reluctant to admit that animals – and therefore machines – are capable of exercising the same level of consciousness that we are.  Brooks assumes that if we continue advancing our robot technology, we will ultimately succeed in creating AI.  To me, though, the true debate is whether or not Artificial Intelligence should really be labeled as “intelligence”.  Searle’s contends that a machine will never really become “intelligent” because the man in the room never truly understands Chinese – he merely follows a set of rules that makes it appear as though he understands Chinese to the people outside of the room.  Brooks makes an important refutation that this statement is ridiculous because each individual neuron inside someone who speaks fluent English does not even understand English. 

Thus, we consider humans intelligent because they are able to produce thoughts as a result of a series of chemical reactions in the brain.  Brooks uses this statement to reason that we should also be able to create a machine whose output feigns intelligence similar to that of a human.  I concur with Brooks on this, however, I have trouble seeing how Chalmers, Penrose, and Searle were so outrageous in asserting that machines lack certain, indefinable “new stuff” that distinguish humans from other species.  Just because science has yet to dub this “new stuff” with a scientific term does not mean that it is complete ludicrous. 

Yet, I believe that it would be a misinterpretation to declare that Searle is arguing that AI is impossible.  Rather, I believe that Searle is arguing that an attempt to reach AI solely through computer programming is foolhardy and doomed to failure.  Instead, we may interpret Searle to suggest that trying to separate the programming from the physical construct in which it is implemented is comparable, and equally incorrect, as attempting to draw a concrete distinction between the mind and the brain.  As he states, “It is best to see strong AI as one of the last gasps of this antiscientific tradition, for it denies that there is anything essentially physical and biological about the human mind.  The mind according to strong AI is independent of the brain.  It is a computer program and as such has no essential connection to any specific hardware.” (Searle, 31)  And this is precisely why Searle believes that the strong AI approach is plainly wrong.  He believes that it is wrong to separate the mind from the very real, physical substance that is the brain.  A computer program cannot replicate the mind because the mind is inextricably connected to the brain.  Thus, I believe that Searle does not deny the possibility that AI can be reached, but does so only on the condition that we acknowledge that AI requires an interaction comparable to the interaction between the brain and what we may more scientifically call consciousness.  A computer version of the human brain is needed.  Searle goes on to put a special emphasis on the uniqueness of the human brain and neuron.  This emphasis is hard to justify, but does not seem ridiculous to assume.

But, it is precisely this emphasis on the special properties of the organic human brain that Brooks contests.  He believes that it might be possible for silicon and steel to make a system that can function in a way similar to that of the brain.  And he presents an argument for it.  Whether or not we subscribe to this argument ultimately depends on our deeply ingrained beliefs about human existence.  Brooks makes no attempt to change these beliefs.

Thus, it is important that when we approach this issue we take Brooks’ point to heart.  We must not let our personal emotions and beliefs about whether or not we are simple machines completely cloud our judgment.  It is impossible for our beliefs not to sway us, but we should be conscious of this and endeavor to fully consider the opposite side.

With that said, I am currently inclined to side with Brooks.  I have no strong reason to believe that human beings are not machines, so I will not deny the possibility that silicon and steel could one-day duplicate human behavior and consciousness.  However, that day has not yet come, so I also see the point of Searle’s conclusion… at least for the moment.

I do not have much knowledge of machines or of the process in which they are built. I am also not hugely religious so my expectation of future technologies is not limited by these beliefs. However I do strongly feel that we are all individuals. The word “machine” to me broadcasts a huge sense of homogeneity and un-uniqueness. Brooks says that we are manufactured by society to have emotions and feelings. Surely this means that we have evolved throughout our lifetime to adapt to certain situations in the ways that current machines could not possibly ever hope to do. It is also my understanding that machines have to be programmed so that their reactions are equal to the stimulus that they are given and so surely this alone sets us apart from machines. We are given advice and rules on how we should behave and react in life however we can equally reject the advice and rules in favor of our own. It seems that there is something inherent in the nature of programming that prevents a machine from deviating from the orders that it is given and even if it does this must too be programmed.

In his critique of Searle he almost reduces his views to ridicule (which he himself criticizes as a form of critique) by using phrases such as “absurdly concludes” and “less sophisticated”. As we discussed in class this puts the reader in a position of severe distrust and doubt. Also as I agreed with Searle in his article and particularly his Chinese room argument I am inclined to disagree with Brooks.

I did however enjoy Brooks’ small entry on how we are special. I had not previously thought of how we classify the respect that we give to different animals in relation to their biological make-up. There is of course no rational reasoning for this but it is in fact human nature. Again I think that human nature is something we cannot re-produce without a sense of programming. Fear is something inherent in humans, not something that is programmed within us and I think it is important to make this distinction. Unless we think of ourselves as programmed from our initial conception I do not believe that Brooks has an argument. This notion is believe is “absurd”.

Don’t get me wrong, I completely believe in the possibility of AI. However, the “artificial” part is what, I suppose, separates me from the ones in the “know.” After our class, Searle’s conclusions were much more evident to me. I originally picked statement number four for his main thesis, but now it is obvious that it is number two. He believes that computers can pass the Turing test but that it is not a valid one to see if a computer can think. I believe that one day all computers will eventually be able to pass the Turing Test but that it does not really reflect anything about consciousness and thinking at all. Thinking, to Searle, is not the mere manipulation of symbols that he describes in the “Chinese room” example. He talks about the importance of the meaning and emphasis that we place on things. But why is Brooks referring to Searle as confused?

Searle and Brooks obviously disagree on the value placed on humanity. Searle ascribes so much more meaning to synapse, neurons, and other impossible to duplicate processes of which Brooks completely disregards. Brooks sees the body as a machine that follows a set of standardized rules and procedure. Agreed. But I believe that there is something in our combination of rules and procedure that allows us at times to go against those same laws (self-preservation vs. suicide).

Brooks also attests to Searle a subconscious fear of alluding to computers as conscious. Searle concludes that “the way that human brains actually produce mental phenomena cannot be solely by virtue of running a computer program” (29). In conclusion, I guess I would never completely cancel out the possibility of creating a “human-like” computer with our technology in the future but at the same time, it will never be an emotive human-being.

Brooks argues that machines are things governed by some set of rules and made up of elements that can be explained and understood with math and science. But he further argues that people are machines because they too can be described by the previous assertion. I agree with Brook’s first point, but I cannot completely agree with his second. He assumes that people act according to a set of rules - all of which are not known to humans, he adds. As far as individual organs and even systems within our bodies, this makes sense, but I don’t buy it when it comes to the human mind. This goes beyond emotions and free will, because the humans are not only capable of making choices. They are capable of acting irrationally, and this simply cannot be swept under the rug of “Brook’s Unknown Rules for Humans”.  

            Brook’s attack on Searle’s arguments, particularly his “Chinese Room”, seems a little misplaced. To me, Searle’s article outlined his technical objections to the Turing test as an appropriate test for a computer’s consciousness. He claims Searle’s use of ridicule is an untenable tactic, but isn’t Brooks doing the same?

Searle did a good job at explaining the weaknesses of the Turing test when it comes to testing whether or not a computer learns and thinks. In the end, I though that Searle’s views were too narrow (probably since the scope of the article was fairly narrow), and I found Brook’ views unsatisfactory and unconvincing. On the question of consciousness, Brook’s relies too much on our continued lack of understanding of the subject. He wants us to do what he has done: accept mediocre, vague responses to difficult questions about our own minds.

After the discussion in class, I’ve become more convinced that questions such as “Is AI possible?” are too vague to without determining what these terms mean. The problem with trying to define these terms, though, is that people’s views about what certain words should mean conflict much, particularly because there is a lot riding on how we use terms. For example, suppose that a computer program were able to pass the Turing test. If we admit that the computer running the program is thinking, then perhaps it follows that it is alive (think Descartes’ cogito ergo sum). Does that mean that turning off the computer is the same thing as killing it? Our intuitions about the answer to these questions that invariably arise prejudice us concerning how we want to use these terms.

So, I have three answers to the question “Is AI possible?” that depend on what AI means. If AI is simply the ability to solve complex problems or perform complex tasks such as playing chess, then AI is certainly possible. If AI is the ability to fool humans that the subject is a computer program, then perhaps AI is possible. If AI has as a necessary component that the subject is alive and has the same moral standing as a person, then AI is not possible.

If one looks at human behavior, one will notice that we are pretty arbitrary in similar, if not identical situations. Now, I question whether or not a computer will have the ability to be arbitrary. Suppose one day that Bob walks into a room covered with red wallpaper and suddenly feels angry; the next day, Bob (his current mood prior to walking into the room is the same as it was during the first day) walks into the same room with red wallpaper and suddenly feels even angrier. Would a computer have this notion of arbitrariness? I tend to think that it would not as it would feel the same level of “angry” after both visits into the room.

My conclusion about AI is: computers someday will have the ability to accurately imitate humans in almost every situation, which will be freaky. There is still a long way to go since humans still do not understand many aspects of themselves that are being imitated.

            Initially, I was skeptical that even if a computer program could pass the Turing test, that artificial intelligence would exist. I think the Searle article did a good job of convincing me that passing of the Turing test didn’t necessarily imply artificial intelligence. Just because a program simulates certain characteristics, it does not mean that the program actually possesses those characteristics. One of his examples that really struck me was the analogy that nobody claims that a program simulating digestion is actually digesting anything, so why would anybody claim that a program simulating intelligence actually has intelligence?

            As usually happens to me when I hear an intelligent debate, though, I was swayed back to the possibility of artificial intelligence by Professor Arora. As he mentioned, if a computer actually could understand the nuances of conversation and fool a human being, how could we definitively say that is does not possess intelligence? This thought resonated with me for a while.

            After much internal deliberation, I’ve decided that I don’t think a computer program will ever really possess intelligence. As Searle mentioned, the human brain is unique in that chemicals affect its behaviors, and the chemicals in a person’s brain (and thus a person’s feelings) effect the way a person holds a conversation. In order for a computer to truly have a conversation the way humans do, and understand things, I think the computer would have to be capable of having emotions based on complex chemical reactions, many of which we don’t even understand in human brains. I really don’t ever see computer programs having true emotions and chemical reactions. Because of this, perhaps one day computer programs can fool us by appearing intelligent, but I personally don’t think they will ever actually BE intelligent, until they get an organ similar to the brain.

I do find Brooks convincing when he argues against the attempt to create centralized systems that organize perception and try to fit it into comprehensive mental maps.  Brooks does persuade me when he claims that the most fruitful direction that AI can take right now is to focus instead on the robot’s or computer’s interaction with the world outside of it, in a more organic and decentralized manner.

Meanwhile, even though Searle might be a little too radical in his argument, there is some nerve that he touches in his article that resonates with things I have been thinking about throughout this class.  One of the issues is related to simulation.  When AI researchers work on computers and robots with the aim of having them pass the Turing Test, of producing artificial intelligence, and in the more ambitious cases of producing consciousness – the way in which they work and their goal is to simulate human intelligence.  The simple fact of simulation raises (for me) a huge doubt as to whether such machines will EVER have real consciousness.  Isn’t simulation by definition supposed to be just that – to come as close to while never reaching the real thing?  Furthermore, and maybe more importantly, if scientists still have not figured it out what it is exactly in the brain that makes us humans conscious, then how can we be so sure that we could ever produce consciousness in anything else?

Even if Brooks is correct in pointing out that once we start building machines that interact with us in such a convincing manner that we start empathizing with them we will have to grant them a level of respect that comes closer to the respect we show other human beings, his view is NOT inconsistent with the fact that these machines, even if we greatly empathize with them, may not have consciousness in the way that we do.  Our own empathy is no measure for an objective fact – it cannot be the answer to the question of whether or not the machines have consciousness.  Brooks fiercely and constantly singles out the emotional desire to maintain our human specialness as the driving force behind arguments against the possibility of producing consciousness in machines.  In order to be consistent, then, he should also factor out any emotional ability we might have to empathize with very advanced robots as a measure for whether they have consciousness.  Neither our desire to be special nor our ability to empathize provide rigorous proof for the possibility of producing conscious machines.  Alas, I am afraid any and all objective proof to support wither side is still missing, so the question remains wide open for speculation.

At this point in the course, after having taken the Touring Test, I realize that as far as we know right now, we cannot create machines that will respond in an indistinguishable manner from humans. However I still am confident in my original view, which relates to that of Searle, that machines are not capable of thinking. Even if machines were able to pass the Touring Test, and could respond in a conversation with humans in a human-like way, the machines are still running on man-made programs. Searle’s Chinese Room argument was perhaps the greatest example for me as to why machines cannot think. Even if machines could carry out human operations such as conversations, they would not be understanding meaning, but would rather only be spitting out symbols in an arranged order. As Searle puts it, machines would be able to understand syntax but not semantics–they could correctly place the language symbols in order so that it would make sense to a human, but they would have no understanding of the meaning of what they had just done.

Rodney Brooks argues conversely that humans are really no different from machines, and in fact could be considered machines themselves. Both are made up of the same molecules that act based on a specific set of rules. Therefore if there is no difference between humans and machines, then machines should be able to think just as the human brain does. I however disagree with Brooks that machines and humans are not fundamentally different, as I think humans, being able to reproduce, or feel pain, or just be alive makes them different from any man-made machine.

Perhaps I could be wrong however, and machines actually are similar to humans in their capabilities for thinking and consciousness.  What is more probably, as Brooks suggests about many other philosophers, is that I am afraid to think that computers could be concious.

Well, I’m scared too, but I must admit that I definitely see where Brooks is coming from.

Is artificial intelligence possible?  After discussion in class, Searle’s argument is much clearer to me.  While he argues that actual “thinking” cannot be elicited from results of a Turing Test, it brings up the question of what would actually show that something is thinking.  I think this is a tremendously difficult question to test because Searle’s definition of thinking incorporates consciousness.  This idea is the greatest problem in the argument about artificial intelligence primarily because it is not understood yet.  The concept of consciousness is not scientific and no one at this point knows the roots of it and what organisms, if any other, possess it.  If we eliminate the consciousness aspect of “thinking,” there is no doubt that artificial intelligence is possible.  Therefore, the main missing crux in the argument for artificial intelligence lies in the possibility of replicating or simulating emotion and consciousness. 

Brooks brings up the question of if something is “functionally” the same as a human, can it be accepted as human?  I will admit that I fall in the group that believes that there is something special about humans.  I am inclined to think that humans encompass a special, inexplicable quality that cannot be duplicated in machines.  While it may not be because of quantum mechanics, as Penrose tried to explain, it still exists, although it eludes us.  If a robot can simulate human behavior, will it actually be human?  I would resolve that it is human-like, but am reluctant to designate it the same level of life as that of a human. 

As history has worked in a way that shows that the unthinkable can be achieved, I am still open minded to the possibility of achieving artificial intelligence.  Although I’m a little skeptical about creating a machine that can function exactly like a human, the speed of technological progress has proved that we should never to rule anything out completely.  I still think that humans have an inherent quality that is unique to us, although I can’t explain where it originates; however, maybe the future will show us differently.