The ID Report

by Denyse O'Leary
ARN correspondent

Here's a longish essay on the state of artificial intelligence, that should be must-reading for people who have been taken in by predictions of thinking, feeling computers.

Not only hasn't it happened, but it is unlikely to happen. Mark Halpern, who has spent fifty years working with computer software helps explain why in the Winter 2006 edition of The New Atlantis.

In October 1950, computer genius Alan Turing published a paper titled "Computer machinery and Intelligence" in the British quarterly Mind, according to which properly programmed computers would generally be accepted as thinking by about 2000. That is, they would successfully respond to human questions in a human-like way. Turing proposed his famous "Turing Test" in that paper. Can you tell the difference between interacting with a computer and interacting with a human?

Anyone who has suffered through the Muzak on a hold line offering a number of useless canned options and responded by hammering the zero button to get hold of a person at a call center in Bangalore, India, or New Brunswick, Canada, can certainly tell the difference. And this is now 2006 and counting.

In 1950 very few people had any idea of the supercomputer-like complexity of a single cell. Your body sheds millions of them every day, shrugs (so to speak), and recruits more. Never mind what the human brain is like either - that's the most complex object in the universe, and it is nothing like a computer. People don't think the way computers process jobs. Never have and never will.

So in the 1950s, large promises were easy to make, and they found their way into abundant science fiction where mechanical brains terrorized the population and warred among themselves. Some, like HAL, were even entertainingly but murderously insane.

According to Halpern, what really happened to the Turing Test was that artificial intelligence (AI) enthusiasts simply changed the original challenge posed by Alvin Turing:

While they have programmed the computer to do things that might have astonished even him, today’s programmers cannot do what he believed they would do—they cannot pass his test. And so the relationship of the AI community to Turing is much like that of adolescents to their parents: abject dependence alternating with embarrassed repudiation. For AI workers, to be able to present themselves as "Turing’s Men" is invaluable; his status is that of a von Neumann, Fermi, or Gell-Mann, just one step below that of immortals like Newton and Einstein. He is the one undoubted genius whose name is associated with the AI project (although his status as a genius is not based on work in AI). The highest award given by the Association for Computing Machinery is the Turing Award, and his concept of the computer as an instantiation of what we now call the Turing Machine is fundamental to all theoretical computer science. When members of the AI community need some illustrious forebear to lend dignity to their position, Turing's name is regularly invoked, and his paper referred to as if holy writ. But when the specifics of that paper are brought up, and when critics ask why the Test has not yet been successfully performed, he is brushed aside as an early and rather unsophisticated enthusiast. His ideas, we are then told, are no longer the foundation of AI work, and his paper may safely be relegated to the shelf where unread classics gather dust, even while we are asked to pay its author the profoundest respect. Turing’s is a name to conjure with, and that is just what most AI workers do with it.

For example, he notes,

In a survey article in the Proceedings of the IRE in 1961, Minsky defends the idea that computers might think by saying that "we cannot assign all the credit to its programmer if the operation of a system comes to reveal structures not recognizable nor anticipated by the programmer," implying that at least some part of such a surprising result must be due to thinking by the machine. He caps his argument with the words: "Turing gives a very knowledgeable discussion of such matters." He quotes nothing specific, just appeals to Turing's stature and authority. But in 2003, Minsky expressed his disappointment and frustration at the lack of progress made by AI toward achieving Turing's goals: "AI has been brain-dead since the 1970s.... For each different kind of problem, the construction of expert systems had to start all over again, because they didn't accumulate common-sense knowledge.... Graduate students are wasting three years of their lives soldering and repairing robots, instead of making them smart. It's really shocking.

One basic problem, as Halpern explains it, is this:

... anyone with an understanding of how computers are made to mimic human responses would need no subject-matter expertise whatever to detect a computer posing as a human. Such a judge would simply demand that the hidden entity respond to the ideas represented by his questions, warning that it would be severely penalized for repeating any of the key words in those questions. Using this interrogative technique, the discriminative ability of judges should increase much faster than programmers' abilities to mimic human responses, and the Test should become correspondingly more difficult and thus more serious.

But in the actual Turing Tests Halpern amusingly describes, the human judges don't appear to make use of any such strategies. One judge decides that a learned scholar is a computer just because she appears to be smart and knowledgeable about her subject.

The 1992 test series Halpern analyzes demonstrate clearly that humans who lack judgement don't make good decisions, but they doesn't demonstrate much else. What people do when they think is just not what computers do.

Toronto-based Canadian journalist Denyse O'Leary (www.designorchance.com) is the author of the multiple award-winning By Design or by Chance? (Augsburg Fortress 2004), an overview of the intelligent design controversy. She was named CBA Canada's Recommended Author of the Year in 2005 and is co-author, with Montreal neuroscientist Mario Beauregard, of the forthcoming The Spiritual Brain (Harper 2007).