Electronic Design

Artificial People: Are They A Fantasy, Or An Ambition

People have wanted to make artificial people for as long as recorded history. In ancient Egypt, priests stood in a hollow statue of a god and bellowed out the god's will. When clockwork mechanisms were developed in the Middle Ages, articulated models of people were made, driven by springs or descending weights. Disney World has lifelike machines acting out scenes from American history, as well as other animatronic displays. Plus, every toy store has mechanical dolls.

The word "robot" was coined by a Czech playwright. Karel Capek, in the early '20s, adapted it from the Czech word for work. The ultimate dream was, and is, a mechanical person who's the slave of a real person. This is pretty easy to accomplish in the imagination, and it has become one of the staple themes of science fiction (one of the greatest oxymorons of all time). It's not quite so easy in real life. It has become useful for commercial prestige to apply the words "robot" and "robotic" to any smart machine. I confess to having done so myself.

With the hype removed, digital computers are electronic clerks that perform data manipulations as instructed by humans. Fantastic clerks, indeed: fast, accurate, tireless, with enormous memories, capable of obeying numerical instructions of vast complexity. They receive data from either human or non-human electrical sources and respond with electric voltages, either digital or analog. The manipulations are called processing, and the instructions are called programs. The responding signals can be displayed on cathode ray tubes, printed on paper, or connected to other machines. Their tremendous utility in science, medicine, engineering, business, war, and personal lives would have been unimaginable, outside of fiction, only a generation ago.

If such clerking is intelligent behavior, then these machines have always been enormously more intelligent than humans. Even mechanical adding machines are much more intelligent than humans by this criterion. I don't pretend to have an adequate description of human intelligence, but it certainly isn't efficient clerking.

Of course, computers compute better than people. There have long been computer programs that can solve complex mathematical equations far better and faster than I can, without making all of my mechanical errors.

But the idea that the human mind is basically a computer and, therefore, bigger and better computers can emulate the human mind is a self-serving act of faith by the computer profession. It's a faith of that profession that software (programming) can do anything, including more programming by itself, given enough hardware (physical computers) and money. After all, DNA is just the program for building people. "Neural networks," "artificial intelligence," "expert systems," and the like do surprising things suggestive of human rational thinking, such as following logic trees in their programs. Some do them better than humans, and some are of great value, but they are not human thinkers and feelers.

If the mind is chemical and physical in the largest sense, we may someday really understand it and synthesize or emulate it. But it isn't clear that more transistors, magnetic spots, and lines of code will do it, no matter how fast and cheap they are. In the real world, we make useful machines, not artificial people. We use computers for all that we can get out of them: we use them in robots and smart machines.

Hide comments


  • Allowed HTML tags: <em> <strong> <blockquote> <br> <p>

Plain text

  • No HTML tags allowed.
  • Web page addresses and e-mail addresses turn into links automatically.
  • Lines and paragraphs break automatically.