The Turing Test and Computer Intelligence

A recent article in The New Atlantis discusses the Turing Test and related issues dealing with intelligence in computers.

Allan Turing was one of the pioneers of computer theory, perhaps the greatest. In 1950, he wrote an article which begins with this sentence: "I PROPOSE to consider the question, 'Can machines think?'" Hardly a modest beginning.

Turing stated that he did not believe in an all-powerful God, or special properties of human immortal souls. Nevertheless, he attempted to consider his question theologically, in particular this objection: "Thinking is a function of man's immortal soul. God has given an immortal soul to every man and woman, but not to any other animal or to machines. Hence no animal or machine can think." He concluded that "In attempting to construct such machines we should not be irreverently usurping His power of creating souls, any more than we are in the procreation of children: rather we are, in either case, instruments of His will providing mansions for the souls that He creates."

As to the objection stated by Turing, I'm not sure it's completely true. I don't find any scriptural evidence that rules out some thinking by animals (or machines, for that matter). Although humans are in the image of God, perhaps other entities (even rocks) have at least a little of God's image in them, too, whatever the image of God is, just as an artist or composer leaves something of herself in all of her creations. And, whatever it is, I don't think that the ability to think is all there is of God's image in humans. I suspect that it includes creativity, the ability to form relationships, the ability to make moral choices, and emotions.

Turing went on to attempt to dispose of eight additional arguments against the possibility of machines being able to think. Perhaps the most important argument, historically, was originally made by Lady Ada Lovelace, also one of the great pioneers of computing, in notes to an article which she translated, and which was published in 1842:

The Analytical Engine has no pretensions whatever to originate anything. It can do whatever we know how to order it to perform. It can follow analysis; but it has no power of anticipating any analytical relations or truths. Its province is to assist us in making available what we are already acquainted with. (from note G) [emphasis is Lovelace's]

Turing said that new types of machine were now available, which hadn't been available in Lovelace's time, which was true, of course.

Toward the end of his article, Turing wrote: "We may hope that machines will eventually compete with men in all purely intellectual fields."

The Turing Test, proposed by Turing, is to communicate with an unknown entity -- carry on a conversation, such as by instant messaging, (which was not available in Turing's day -- back then some sort of typing would have been used) and see if you can tell if the unknown is a human or a computer. Turing said that, if we cannot be sure, then the computer has achieved intelligence.

The article in The New Atlantis basically says that Lady Lovelace was right, at least so far. The Turing Test criterion, says the author, has not been met. Computers, so far, seem only able to do what they have been programmed to do. Mark Halpern, the author, concludes that ". . . we have not yet achieved artificial intelligence, and have no idea if we ever will." This in spite of Turing's affirmation that we will achieve this. Halpern discusses some attempts to apply the Turing Test, and related matters.

Note: The New Atlantis article is readable by any reasonably intelligent person. The article by Turing is also mostly readable, but considerably longer. The article translated by Lovelace, and most of her notes, are not for the faint-hearted.

Note: This post was revised somewhat on March 13th. Thanks for reading!