Cognitive science: The glad, sad, mad computer - or anyway, merry chrysanthemum!

As Joseph Dumit (STS, University of California Davis) will tell it tomorrow at York University tomorrow (Tuesday, 24 March 2009, 12h30 - 14h0), Paul A. Delaney Gallery (320 Bethune College):

True Demons of Cognition: When Computers were Glad, Sad, and Mad yet Logical; or, a Brief History of Experimental Epistemology at the End of Cognitive Science

This talk proposes that we are approaching the limits of a cognitive neuroscience approach that adopts a too-simple model of circuits to account for emotions, pathology, meditation, and subjectivity. During the 1950s and ‘60s, experimentation with circuits co-produced computers, cognitive psychology, cybernetics, psychiatry, anthropology and psychoanalysis, with sometimes disturbing results. This work valued circuits for their innate, pathological irrationality, and for the way in which they portrayed time (subjective, logical, existential and even psychoanalytic) as something that fed back into a better understanding of the wiliness of machines. Such conceptions of time, however, proved short-lived. They were erased, curiously, by the advent of a cognitive psychology that took computers as models of (human) rationality, and which subsequently generated an image of computers as taken-for-granted objects, completely understood.

I am not sure what any of this means. The human brain is nothing like a computer, and efforts to make it appear so are efforts at creating illusions. I have long thought that the brain is more like an ocean, in the sense that you can drop down a probe, and turn up something you never expected. And every brain is different, so there are billions of oceans out there. A research heaven, if only the researcher were immortal.

Merry Chysanthemum? Oh, that's the computer's first Christmas card. Gives you some idea what to expect.

See also:

Mind: Yet another effort to explain to materialists why minds are not like computers; Getting computers to pretend to converse is an extremely hard computational problem; Artificial intelligence: Conversing with computers or with their programmers?; Computers: Most engineers must have that they themselves are not robots; Artificial intelligence: A look at things that neither we nor our computers can discover; Can a conscious mind be built out of software?; Mind vs. meat vs. computers - the differences; Let the machine read your mind (We offer an installment plan!); Mind-computer blend: Who believes in this?; Artificial intelligence: Making the whole universe intelligent?; Brain cells release information more widely than previously thought.

(Note: Thanks much to kind PayPal donors. As media move online, it is best to support the news you want to see. - Denyse)

Hat tip: The Sheepcat

Mind: Yet another effort to explain to materialists why minds are not like computers

In "Why minds are not like computers," in The New Atlantis (Winter 2009) Ari N. Schulman explains:

Hmmm. What can go wrong with that reminds me of "The Computer's First Christmas Card," "Merry Chrysanthemum". And as a result,

Thus it is only partially correct to say that a computer performs arithmetic calculations. As a physical object, the computer does no such thing—no more than a ball performs physics calculations when you drop it. It is only when we consider the computer through the symbolic system of arithmetic, and the way we have encoded it in the computer, that we can say it performs arithmetic.

Okay, but what about the famed "Turing test" that would show that computers can think?

... they indicate crucial errors in AI researchers’ understanding of both computers and minds. Suppose that the mind is in fact a computer program. Would it then be possible to conclude that what’s inside the mind is irrelevant, as is supposed by some interpretations of the Turing Test? If we have some computer program whose behavior can be completely described as if it were a black box, such a description does not mean that the box is empty, so to speak. The program must still contain some internal structures and properties. They may not be necessary for understanding the program’s external behavior, but they still exist. So even if we possessed a correct account of human mental processes in purely input-output terms (which we do not), such an external description by definition could not describe first-person experience. The Turing Test is not a definition of thinking, but an admission of ignorance—an admission that it is impossible to ever empirically verify the consciousness of any being but yourself. It is only ever possible to gain some level of confidence that another being is thinking or intelligent. So we are stuck measuring correlates of thinking and intelligence, and the Turing Test provides a standard for measuring one type of correlate. Similarly, although social interaction requires communication in the form of such “input-output” as speech and hearing, it also requires two or more agents who experience that interaction: A teddy bear may provide a child with the same comfort and companionship—and elicit from the child the same responses—that another human being would, but we cannot say that it loves.

Schulman, whose comments I commend to all, is assistant editor of The New Atlantis.


Personally, I think that the artificial intelligence project is dead, but no one has got round to raising the funds for a decent burial.

See also:

Getting computers to pretend to converse is an extremely hard computatinal problem

Artificial intelligence: Conversing with computers or with their programmers?

Computers: Most engineers must have that they are not robotsArtificial intelligence: A look at things that neither we nor computers can discover

Can a conscious mind be built out of software?

Also: Mind vs. meat vs. computers - the differences

Let the machine read your mind (We offer an installment plan!)

Mind-computer blend: Who believes in this? Artificial intelligence:

Making the whole universe intelligent?

Brain cells release information more widely than previously thought.

(Note: Thanks much for all kind readers who are still here. I would have hoped to be here more often too. And thanks to kind PayPal donors. As media move online, it is best to support the news you want to see. - Denyse)

Mind vs. meat vs. computers - the differences

Some thoughts from Roy Abraham Varghese, co-author of There IS a God:

Why our minds are not meat:

"Beyond consciousness, there is the phenomenon of thought, of understanding, seeing meaning. Every use of language reveals an order of being that is innately intangible. At the foundation of all of our thinking, communicating and use of language is a miraculous power. It is the power of noting differences and similarities and of generalizing and universalizing- what the philosophers call concepts, universals, and the like. It is natural to humans, unique, and simply mystifying. How is it that, from childhood, you can effortlessly think of both your dog Caesar and dogs in general? You can think of redness without thinking of a specific red thing (of course redness does not exist independently, but only in red things). You abstract and distinguish and unify without giving your ability to do these things a second though. And you even ponder things that have no physical characteristics, such as the idea of liberty or the activity of angels. This power of thinking in concepts is by its very nature something that transcends matter." (There Is A God, pp. 176-177)

Make no mistake, this problem - called the problem of "qualia" by philosophers of mind - is a big, unsolved, unsolvable* problem for materialist theories. By the way, for fun, share the aliens' amazement at the way, on this planet, "meat" can think

Why our mind/brain complex is not a computer:

Many misconceptions about the nature of thought arise from misconceptions about computers. But let's say we're dealing with a supercomputer like the Blue Gene, which does over two hundred trillion calculations per second. Our first mistake is to assume that Blue Gene is an "it" like a bacterium or a bumblebee. In the case of the bacterium or the bumblebee we're dealing with an agent, a center of action that is an organically unified whole, an organism. All its actions are driven by the goals of maintaining itself in existence and replicating. Blue Gene is a bundle of parts that jointly or severally perform functions "implanted" and directed by the creators of the collection. Second, the bundle of parts does not know what "it" is doing when "it" performs a transaction. Supercomputer calculations and mainframe transactions performed in response to data and instructions are purely and simply a matter of electrical pulses, circuitry, and transistors. The same calculations and transactions performed by a human person, of course, involve the machinery of the brain, but they are performed by a center of consciousness that is conscious of what is going on, understands what is being done, and intentionally performs them." (There Is A God pp 178-179)

Sometimes, articles in popular media are written in a misleading way, so that we are led, for example, to assume that a computer is playing a grandmaster at chess. In fact, a programmer has programmed the computer, in anticipation of the grandmaster's moves. The computer is not playing the grandmaster; the programmer is - working through a software design that the programmer hopes will cover enough bases to win. Sometimes the programmer wins and sometimes he doesn't. As Mario and I say in The Spiritual Brain,

... a powerful computer program is capable of considering many more strategies at once than a human being can. Generally, a chessplaying computer relies on its enormous parallel processing power to sort through a vast memory to evaluate millions of moves and choose the best one. Deep Junior powered through up to 3 million possible moves per second. Kasparov probably evaluated only two to three moves per second. Well, that raises an obvious question: Why does Kasparov ever win? Shouldn’t he always lose?

The answer seems to be that what Kasparov is doing when he is thinking about his next play is different in kind from what Deep Junior is doing. Kasparov himself said, “Whatever [programmers] Shay and Amir say about Junior’s ability to run through millions of possible strategies, I, by contrast, might consider only a few strategies in any one game. But you can bet your life they will be the very best ones.” As philosopher and chess enthusiast Tim McGrew, of Western Michigan University, puts it: “Something is going on in the grand master’s mind that is not only radically different . . . but also inconceivably more efficient. It is a kind of computational miracle that humans can play chess at all.” (P. 22)

*Problems become unsolvable when the operative assumptions are wrong, but are insisted on anyway.