Are We Unique?: A Scientist Explores the Unparalleled Intelligence of the Human Mind

Rating:
Summary: disappointing
Review: From a spiritually-troubling encounter with AI to the motive for his argument - "I am deeply troubled by the notion that all of the great achievements of our species...are nothing more than the results of...machines that we all carry around in our skulls", the premises of his argument are unsubstantiated and unscientific, reflecting the rest of the book.

His lack of reason can be seen in arguments such as that the brain is not a computer because it operates on a mechanical level at a much slower speed; that though computers can resemble the brain in some ways, they have to do this through such methods as parallel processing, which is "not natural" for the computer; that the brain evolved, and the computer designed shows they must be different. He also got Moore's Law wrong.

If you are interested in the subject, I would recommend a much more pleasurable and scientifically-based read such as "How the Mind Works" by Steven Pinker.

Rating:
Summary: Trefil proves Humans are more human than any other species.
Review: I have read a number of Trefil's books, including "Meditations at 10,000 feet: A scientist in the mountains," and "From atoms to quarks." Ordinarily, I enjoy Trefil's books, which are typically written with a clear and concise style that is at once entertaining, informative, and easy to read. It was with some disappointment, therefore, that I read his latest book "Are we unique." Trefil seems to have left some of his scientific objectivity behind in a personal quest to show humans are not only trivially unique, but somehow the most unique thing in the universe.

To his credit, Trefil openly admits his personal objective, where he says:

Make no mistake, though. This isn't going to be a cool, dispassionate examination of an intellectual problem. I desperately want to find a way out of this dilemma [about human uniqueness], and intend to devote whatever scientific skills I've developed in my career to finding it." [See page 13.]

The first problem is that Trefil never really defines what he means by uniqueness. Throughout the book, I alternately got the impression that uniqueness to him means not reproducible, different in magnitude from all other species, or more complex and wonderful than all other things. The problem is that he never offers a good explanation why any of this is either true, or important. For example, humans are clearly reproducible. All you have to do is visit the local maternity ward to prove that. Yet, Trefil spends almost all of the second half of the book arguing that a computer cannot reproduce human thought, without clearly explaining why silicon is the issue. What if humans derive the future ability to build a more intelligent biological lifeform? What would that do to his premise?

Trefil argues that simply because other species have the germ of human capabilities, such as limited ability to communicate, humans are so far advanced that we deserve unique status based on magnitude alone. What continually strikes me about this argument is the amount of specism inherent in the tests. For example, Trefil shows how lobsters use their sense of smell to discern the presence of other lobsters by the release of urine in the water. Essentially, lobsters "see" through their sense of smell. Then, he proceeds to describe experiments in which animals are shown their image in a mirror, with a bright red spot painted on their forehead. Those who recognize their image, and try to get the red spot off, are categorized as "self aware," while those who blithely ignore the image receive a lesser classification. I could not help wondering if an intelligent race of lobsters would deem humans not self aware because, upon being tossed into a swimming pool with their urine, they failed to recognize themselves. We are visual animals, and place an inordinate amount of emphasis on that sense, forgetting that other animals see the world through senses we hardly recognize in our daily lives.

The other part of this argument I find very illuminating is that humans invariably focus on things we consider to be good qualities when asserting our special uniqueness among animals. Trefil focuses on intelligence, for example. He might just as well have argued that humans are unique because we are the only species to incarcerate other members of our clan. This tendency to find uniqueness in good traits, while ignoring our dark, but equally unique nature, makes me wonder if the entire discussion isn't so contaminated by the bias of specism as to make the conversation pointless.

Having demonstrated, to his satisfaction, that humans are unique among animals, Trefil proceeds to show that no computer can mimic the human mind. Here, I find he makes some rather good points. There really is no reason to assume a digital computer can model the human mind, since there is no good reason to assume the human mind is digital in the first place. I particularly enjoyed his discussion about how the human brain works. In making his point, Trefil touches on chaos, the electro-chemical nature of the brain, self-organization, and emergent properties of complex systems. All of this makes for interesting reading.

Throughout the book, I found myself asking why this uniqueness seems so important to some individuals. Personally, I have never felt threatened by a computer that could beat me in chess (an easy thing to do), a car that goes faster than I can run, or an animal that might learn to communicate with me. I am not at all afraid of discovering my biological connection with animals. Consequently, I had a consistently difficult time relating to the concerns that motivated Trefil to write his book.

Just the same, Trefil's book was worth reading, and I found much of the supporting information particularly insightful. I am not sure he supported his thesis very well, and I do not understand the motivation to assert human uniqueness (at least in the sense that we are somehow more unique than other species). Just the same, it is good to read alternate points of view. This is a complicated area, and anyone interested in where we fit in the cosmos should hear all sides. Other books related to this subject matter include "When Elephants Weep," by Jeffrey Moussaieff Masson and Susan McCarthy, and "Shadows of Forgotten Ancestors," by Carl Sagan and Ann Druyan.

Duwayne Anderson

Rating:
Summary: Trefil proves Humans are more human than any other species.
Review: I have read a number of Trefil's books, including "Meditations at 10,000 feet: A scientist in the mountains," and "From atoms to quarks." Ordinarily, I enjoy Trefil's books, which are typically written with a clear and concise style that is at once entertaining, informative, and easy to read. It was with some disappointment, therefore, that I read his latest book "Are we unique." Trefil seems to have left some of his scientific objectivity behind in a personal quest to show humans are not only trivially unique, but somehow the most unique thing in the universe.

To his credit, Trefil openly admits his personal objective, where he says:

Make no mistake, though. This isn't going to be a cool, dispassionate examination of an intellectual problem. I desperately want to find a way out of this dilemma [about human uniqueness], and intend to devote whatever scientific skills I've developed in my career to finding it." [See page 13.]

The first problem is that Trefil never really defines what he means by uniqueness. Throughout the book, I alternately got the impression that uniqueness to him means not reproducible, different in magnitude from all other species, or more complex and wonderful than all other things. The problem is that he never offers a good explanation why any of this is either true, or important. For example, humans are clearly reproducible. All you have to do is visit the local maternity ward to prove that. Yet, Trefil spends almost all of the second half of the book arguing that a computer cannot reproduce human thought, without clearly explaining why silicon is the issue. What if humans derive the future ability to build a more intelligent biological lifeform? What would that do to his premise?

Trefil argues that simply because other species have the germ of human capabilities, such as limited ability to communicate, humans are so far advanced that we deserve unique status based on magnitude alone. What continually strikes me about this argument is the amount of specism inherent in the tests. For example, Trefil shows how lobsters use their sense of smell to discern the presence of other lobsters by the release of urine in the water. Essentially, lobsters "see" through their sense of smell. Then, he proceeds to describe experiments in which animals are shown their image in a mirror, with a bright red spot painted on their forehead. Those who recognize their image, and try to get the red spot off, are categorized as "self aware," while those who blithely ignore the image receive a lesser classification. I could not help wondering if an intelligent race of lobsters would deem humans not self aware because, upon being tossed into a swimming pool with their urine, they failed to recognize themselves. We are visual animals, and place an inordinate amount of emphasis on that sense, forgetting that other animals see the world through senses we hardly recognize in our daily lives.

The other part of this argument I find very illuminating is that humans invariably focus on things we consider to be good qualities when asserting our special uniqueness among animals. Trefil focuses on intelligence, for example. He might just as well have argued that humans are unique because we are the only species to incarcerate other members of our clan. This tendency to find uniqueness in good traits, while ignoring our dark, but equally unique nature, makes me wonder if the entire discussion isn't so contaminated by the bias of specism as to make the conversation pointless.

Having demonstrated, to his satisfaction, that humans are unique among animals, Trefil proceeds to show that no computer can mimic the human mind. Here, I find he makes some rather good points. There really is no reason to assume a digital computer can model the human mind, since there is no good reason to assume the human mind is digital in the first place. I particularly enjoyed his discussion about how the human brain works. In making his point, Trefil touches on chaos, the electro-chemical nature of the brain, self-organization, and emergent properties of complex systems. All of this makes for interesting reading.

Throughout the book, I found myself asking why this uniqueness seems so important to some individuals. Personally, I have never felt threatened by a computer that could beat me in chess (an easy thing to do), a car that goes faster than I can run, or an animal that might learn to communicate with me. I am not at all afraid of discovering my biological connection with animals. Consequently, I had a consistently difficult time relating to the concerns that motivated Trefil to write his book.

Just the same, Trefil's book was worth reading, and I found much of the supporting information particularly insightful. I am not sure he supported his thesis very well, and I do not understand the motivation to assert human uniqueness (at least in the sense that we are somehow more unique than other species). Just the same, it is good to read alternate points of view. This is a complicated area, and anyone interested in where we fit in the cosmos should hear all sides. Other books related to this subject matter include "When Elephants Weep," by Jeffrey Moussaieff Masson and Susan McCarthy, and "Shadows of Forgotten Ancestors," by Carl Sagan and Ann Druyan.

Duwayne Anderson

Rating:
Summary: disappointing
Review: Trefil has a long history of "popular science" books for the laymen which minimize, or outright misrepresent, current scientific debate. His latest is no exception.

"Are We Unique?" starts with a preface which states the author's bias in an unusually revealing manner. After experiencing a computer "artificial intelligence" program designed to find patterns, Trefil was unnerved by the result: "I was listening to the words of an intelligence that was totally alien." On leaving the AI lab, Trefil makes a big point of the fact that he crossed himself.

This, in a nutshell, is the book's problem. Trefil's twin thrusts are: 1. If it is "alien" it cannot be human, and if it is not human it cannot be intelligence, and; 2. Christianity (although never mentioned by name) underscores all claims to human uniqueness.

Trefil uses classic obfuscationist techniques such as bait-and-switch and "everybody-knows" dismiss! als to completely misrepresent the issues. On the subject of cetacean intelligence, Trefil dismisses the entire field with a breezy (and undocumented) "scientists no longer accept the notion that dolphins are in some way more intelligent than other animals." On the subject of the ape language experiments, Trefil again dismisses the entire field out-of-hand: "The consensus among scientists these days seems to be that those early claims for language abilities in great apes cannot be substantiated." (Needless to say, the *scientists* involved in these studies would probably be surprised to hear of this!) When it comes to computer Artificial Intelligence, Trefil is reduced to saying, "We can't really make progress in our quest for human uniqueness unless we dispose of this particular canard [that the brain is just a computer]."

In other words, the book is not an investigation of whether human intelligence is unique, but a "quest" to ! prove it is so -- against mounting evidence to the contrary! .

And why? Because Trefil's Christian beliefs make him unable or unwilling to accept that other "alien" intelligences do (or soon will) exist. He clings to his "uniqueness" with all of the desperation of the truly scared.

He SHOULD cross himself.

Rating:
Summary: A Scientist Experiences a Crisis of Faith
Review: Trefil has written some good introductory general science books, but this isn't one of them.

The book sets out to answer the question posed in the the title. It's OK that he wants to answer it positively. (Apparently, he feels that he NEEDS to answer it positively.) It's OK that he presents his arguments in a way that is intended to make them convincing. But it is not OK that he falls into a very serious intellectual trap.

Exactly at the moment where the problem gets most interesting (and difficult), Trefil takes the easy way out and appeals to common sense and intuitive understandings. To quote Steven Brust, "simple things are never problems". And yet, Trefil seems to want to find solutions that require the problems involved to be simple.

For instance, he spends a whole chapter on Searle's "Chinese Room" thought experiment. Ultimately, the Chinese Room is a trick. Searle's argument turns on the question of what it is to "know" Chinese. Is it enough to speak Chinese fluently? Do you really "know" Chinese if people ask you questions in Chinese and you provide them correct answers? Most people would say "yes". But Searle's thought experiment slows that down and imagines a little man in the head who doesn't know Chinese, but just has a set of rules for listening to one set of sounds and responding with another. "Obviously" this mechanical version of the mind does not "understand" Chinese, but would pass a Turing test.

Trefil takes this to be a serious refutation of the computational theories of the mind, but only by not understanding the rebuttals. When it is pointed out that it would be impossible for a mechanical set of rules to actually speak fluent Chinese, he says "it was just a thought experiment so it doesn't have to be possible." And yet, that is the entire issue! If "knowing" Chinese can be extracted from the mind and reproduced in a set of mechanical rules, then is not the mind really mechanical after all? And if it can't be done, then the "thought experiment" is actually intellectually sterile.

Again and again Trefil performs the same "bait and switch" in this book - setting up a discussion that is supposed to be grounded in science but which slips into fuzzy mysticism at just the moment where the science actually needs to be most grounded.

I doubt he would let any of his physics students pull the same stunt in the answers to their assignments.

Rating:
Summary: Is this a case of "true, but trivial?"
Review: What Trefil demonstrates here is that we are indeed unique and always will be regardless of how powerful and clever and "conscious" our computers become. He is NOT that we will be unsurpassed in intelligence or that we possess souls and computers don't. What Trefil is asserting is something much more modest. We are unique in the sense that there is nothing exactly like us. This is a "unique" akin to the "special" in "special education." We are special. It doesn't matter how filled or unfilled the universe is with intelligent life forms. We are and will be uniquely ourselves.

Trefil's main subsidiary argument, that the difference between animal consciousness and human consciousness is more than just a difference of degree, is more interesting. He asserts that the difference is so great that it constitutes a difference of kind. When things get sufficiently complex we have the "emergent property" phenomenon. As others have said, more is different, and what emerges is something that cannot always be predicted.

I also think he makes a nice argument for the possibility that we will not be able to build a computer more sophisticated than our brain. It could just be, he asserts, that something like Godel's theorem which states that any mathematical system of sufficient complexity includes statements that can't be proven or are contradictory, may be "waiting for us in complex systems" (p. 219) His point is there may be a practical barrier beyond which our very finite minds cannot go. I suspect he is right, and this is something worth pondering.

But there is a problem with the intent of this book. Notice first that the title refers to the "unparalleled" intelligence of the human mind, not the "unsurpassed." One could also say that the intelligence of e.g., a silverfish is "unparalleled." Trefil knows that our intelligence is very likely to be surpassed. This sly use of language is similar to his main assertion that we are "unique." It's true, but so what?

Perhaps Trefil set out to show that human consciousness is something that a computer could never achieve. To this end he surveyed the "landscape of opinion" on consciousness and found the following categories: "Deniers," those who deny that there is a "problem of consciousness at all-that once you understand what the neurons are doing, there's nothing else to explain" (pp. 182-183). He puts Daniel Dennett in this category. Next there are the "Mysterians," those who feel that the problem of consciousness will never be solved. And finally there are the "Materialists," like Francis Crick, who in his book, The Astonishing Hypothesis (1994), asserts that we are "in fact no more than the behavior of a vast assembly of nerve cells and the associated molecules."

Trefil (who is a physicist) claims to be a member of this latter group. However I think that he has a lot of the "mysterian" in him. For example, he writes on page 173 that when he was courting his future wife he suddenly realized he loved her, and he realized he knew this with more certainty than he had ever known anything in physics or mathematics. He adds that an algorithm running on a Turing machine is never going to know anything in quite the same way.

I agree, but the knowing he is talking about is hardly just a knowing. It is a near total brain/body experience, and since he is an evolutionary, reproducing human being, it's a profound and extremely important experience. He isn't just "knowing." He is feeling and then some. Computers don't feel. So of course they cannot know something in quite the same way.

There are other (small) problems with Trefil's presentation, but I'll skip them and go straight to what I think is the most glaring: an omission of a category for those who believe that consciousness is simply an illusion, a device of the evolutionary mechanism. This category would include most evolutionary psychologists and Tor Norretranders, who wrote The User Illusion: Cutting Consciousness Down to Size (1991, 1998), a book I wish Trefil had read. From this point of view our consciousness is an emergent property of the evolutionary process, a mechanism that forces us to identify with the particular phenotype that we are and to work tirelessly for its advantage and to fear its disadvantage, especially its death. A computer, which has not gone through the same sort of process would not have a similar consciousness. On pages 184-185 Trefil encounters this point of view from the "Illusionists" (as I'll dub them), but misses the significance of what they said to him. He writes that the neuroscientists with whom he was talking were "apt to brush off questions about consciousness with a wave of the hand and, ‛Oh, it's just an illusion,' then get back to their work." Trefils claims these people are "deniers," not realizing that there is a world of difference between calling something an illusion and saying it doesn't exist. Illusions exist, and can be very powerful. Our consciousness is an illusion we can't help but believe.

I'd like to add that this idea of consciousness is an old one that can be found in the works of Buddhism, jnana yoga and the Vedas, and probably elsewhere. Our consciousness is an amazing experience, so overwhelming and so overpowering that it dominates all aspects of our existence. But it's just part of the illusion, the maya that separates us from the underlying truth of our existence. Our ego, our subjective sense of self, is also an illusion. Our individual consciousness is no different than the other person's. It is exactly the same: a mechanism of profound identification that fosters the delusion of separateness.

Rating:
Summary: Is this a case of "true, but trivial?"
Review: What Trefil demonstrates here is that we are indeed unique and always will be regardless of how powerful and clever and "conscious" our computers become. He is NOT that we will be unsurpassed in intelligence or that we possess souls and computers don't. What Trefil is asserting is something much more modest. We are unique in the sense that there is nothing exactly like us. This is a "unique" akin to the "special" in "special education." We are special. It doesn't matter how filled or unfilled the universe is with intelligent life forms. We are and will be uniquely ourselves.

Trefil's main subsidiary argument, that the difference between animal consciousness and human consciousness is more than just a difference of degree, is more interesting. He asserts that the difference is so great that it constitutes a difference of kind. When things get sufficiently complex we have the "emergent property" phenomenon. As others have said, more is different, and what emerges is something that cannot always be predicted.

I also think he makes a nice argument for the possibility that we will not be able to build a computer more sophisticated than our brain. It could just be, he asserts, that something like Godel's theorem which states that any mathematical system of sufficient complexity includes statements that can't be proven or are contradictory, may be "waiting for us in complex systems" (p. 219) His point is there may be a practical barrier beyond which our very finite minds cannot go. I suspect he is right, and this is something worth pondering.

But there is a problem with the intent of this book. Notice first that the title refers to the "unparalleled" intelligence of the human mind, not the "unsurpassed." One could also say that the intelligence of e.g., a silverfish is "unparalleled." Trefil knows that our intelligence is very likely to be surpassed. This sly use of language is similar to his main assertion that we are "unique." It's true, but so what?

Perhaps Trefil set out to show that human consciousness is something that a computer could never achieve. To this end he surveyed the "landscape of opinion" on consciousness and found the following categories: "Deniers," those who deny that there is a "problem of consciousness at all-that once you understand what the neurons are doing, there's nothing else to explain" (pp. 182-183). He puts Daniel Dennett in this category. Next there are the "Mysterians," those who feel that the problem of consciousness will never be solved. And finally there are the "Materialists," like Francis Crick, who in his book, The Astonishing Hypothesis (1994), asserts that we are "in fact no more than the behavior of a vast assembly of nerve cells and the associated molecules."

Trefil (who is a physicist) claims to be a member of this latter group. However I think that he has a lot of the "mysterian" in him. For example, he writes on page 173 that when he was courting his future wife he suddenly realized he loved her, and he realized he knew this with more certainty than he had ever known anything in physics or mathematics. He adds that an algorithm running on a Turing machine is never going to know anything in quite the same way.

I agree, but the knowing he is talking about is hardly just a knowing. It is a near total brain/body experience, and since he is an evolutionary, reproducing human being, it's a profound and extremely important experience. He isn't just "knowing." He is feeling and then some. Computers don't feel. So of course they cannot know something in quite the same way.

There are other (small) problems with Trefil's presentation, but I'll skip them and go straight to what I think is the most glaring: an omission of a category for those who believe that consciousness is simply an illusion, a device of the evolutionary mechanism. This category would include most evolutionary psychologists and Tor Norretranders, who wrote The User Illusion: Cutting Consciousness Down to Size (1991, 1998), a book I wish Trefil had read. From this point of view our consciousness is an emergent property of the evolutionary process, a mechanism that forces us to identify with the particular phenotype that we are and to work tirelessly for its advantage and to fear its disadvantage, especially its death. A computer, which has not gone through the same sort of process would not have a similar consciousness. On pages 184-185 Trefil encounters this point of view from the "Illusionists" (as I'll dub them), but misses the significance of what they said to him. He writes that the neuroscientists with whom he was talking were "apt to brush off questions about consciousness with a wave of the hand and, ‛Oh, it's just an illusion,' then get back to their work." Trefils claims these people are "deniers," not realizing that there is a world of difference between calling something an illusion and saying it doesn't exist. Illusions exist, and can be very powerful. Our consciousness is an illusion we can't help but believe.

I'd like to add that this idea of consciousness is an old one that can be found in the works of Buddhism, jnana yoga and the Vedas, and probably elsewhere. Our consciousness is an amazing experience, so overwhelming and so overpowering that it dominates all aspects of our existence. But it's just part of the illusion, the maya that separates us from the underlying truth of our existence. Our ego, our subjective sense of self, is also an illusion. Our individual consciousness is no different than the other person's. It is exactly the same: a mechanism of profound identification that fosters the delusion of separateness.