Alan Turing, the Twentieth Century British mathematician and Philosopher, designed the early computer that broke the Nazi code, saving D-Day, for example.
He designed a test to determine whether one was communicating with a computer or a human. This test delights us still today and is the basis of much of has become of contemporary Metaphysics, "How do we know the Self?"
He had based his computing system on logic, and since humans are, by default, based on logic, they should be similar. But computers are not wired for emotion. Can they mimic?
His test was whether or not the person on one side of the communication was fooled that the "person" on the other side was, in fact, not a human at all, but a computer, a software computing system based on the one's and zero's of the German Rationalist, Leibnitz.
In addition to the binary number system, back in the 1600's, Leibnitz had also developed calculus. Where would we be without that? He was, in addition the founder of symbolic logic, which, as I said, Turing used to design the underlying principles of the early computer.
The thought experiment that Turing laid out involved a wall with desks and chairs in front of a little slot, like a letter slot, through which notes could be passed back and forth. Or, in the technology of the day, a teletype machine could be used. How would we know if the responses from the other side came from a human or from a machine?
Since the popularity of electronic mail, we are all able to imagine this experiment. "Was the email generated by a "spam" computer?" is a different question than, "Was the email written by a spammer?" In other words, were the words chosen by a human, or by software, or by software designed by a human?
We are all familiar now with the electronic voice, generated by computer program, that directs us through a telephonic menu. We recognize the voice as the recorded voice of a human or a computer-generated voice, but know that whatever it is, it is not a full human. Instead, these are bits tied together to fulfill a function.
Humans are more than functionaries. Yet, we all have known someone that acted so mechanically or had such a monotone of emotion, that we doubted if that person was fully human. They seem to have failed the Turing test!
We delight in playing with computer games that appear to synthesize humanness. Yet, humanity is more than a puzzle to be assembled from parts. Unlike geometry, the property of being human is more than the sum of the parts. Yet, how do we know?
Is "artificial intelligence" sufficient to design a pretend human? Where are the emotions? Even the taped voices on the telephone menus now often have a sing-song melody to imitate true emotion. But what about those people who do not "express" emotion? How can pretend emotion from a computer-generated synthesizer be confused with a non-emotional human? Finally, can computers someday be programmed to have emotion?
I would contend that it is more than reason and emotion that compose the human interiority; intentionality is also a basic element. One thing that defines any human tool or creation is its lack of will. It is used, rather than being a user. While a robot can be programmed to use a wrench, it cannot be programmed to intend to do anything: it can only follow instructions.
As long as artificial intelligence is limited by its being a tool rather than an existential interiority, capable of feeling unfulfilled or meaninglessness, then it is not fully human. Without doubt and longing, how can the computer be human? Can a computer be kind? Can a computer be a friend? Yet, some humans are neither.
On the other hand, we can project so many feelings upon an inanimate object that we "believe" a toy is a person. Stuffed animals "offer" undying affection. In regard to the animate world, many people think of their pets as their best friends. Surely, there is a special kind of animal responsiveness that leads to an attribution of "intelligence." Many gardeners talk to their plants and claim it makes a difference in growth: "sweet talk" works, but angry words inhibit growth. Some mechanics say that "sweet talk" to a car makes it run better.
We love to project human potential onto that which inanimate, neither plant nor animal. Like the ancient Greek story of the sculptor whose statue came to life, we like to think our human creations can become "real."
The study of artificial intelligence entertains us, but re-creating humanity eludes us. By definition, humans are not artifical creations.
The Turing test motivates Cognitive Science and delights us all.
No comments:
Post a Comment