Newton Blog

« How I Used Math to Beat a Speed Camera Ticket | Blog Home Page | We Made Portland Angry. We Don't Apologize »

I'm Sorry Dave, I'm Afraid I Can't Do That

Hal 9000 is perhaps the best and most iconic incarnation of the intelligent machine in science fiction. Rosey, R2-D2, Marvin the Paranoid Android, Skynet, V.I.C.K.I., The Borg, The Matrix, WALL-E -- robots and computers of the future bring human-level intelligence to bear on tasks ranging from doing the laundry to dominating the planet. (Or just being morose.)

KITT: The artificially intelligent Trans-Am.

Science fiction is the art of melding writing and film with the concepts of science, or at least with imagery derived from science. The artistic element of artificially intelligent (AI) machines is a proven winner at the box office and on the shelves. But, when will we actually produce a computer that rivals a human? History and the predictions of the past can give some insight.

It didn't take long, after the design of the first computational machines, for their inventors to consider turning the virtual gaze of their creations from code-cracking and algebra-solving to human intelligence. One of these inventors was Alan Turing, a British genius who created the famous Turing test. A computer and a real person both speak with a judge who doesn't know which is which. If the judge cannot succesfully distinguish the computer from the human more often than not, the computer passes the test.

In 1950, Turing predicted that a computer would pass the test by 2000. His foresight regarding how powerful future computers would become was eerily accurate. However, if you've ever had a characteristically bizarre, obtuse and watery thin conversation with a "chat robot", you're familiar with the continuing failure of his AI prediction.

Early progress in replicating human intelligence was rapid. Machines that could carry out behaviors such as recognizing and stacking wooden blocks were developed within two decades of the arrival of the first computers. A mood of euphoria swept through the field, and researchers felt confident about the future. Marvin Minsky, a founder and giant in the area of AI research, predicted that we would have artificial intelligence within a generation. Since this prediction was issued in 1967, we can call it a bust.

Many notable AI researchers, as well as psychologists, writers and others who studied the progress of computer intelligence throughout the 1960s and 1970s believed that AI was only 15-35 years away. As these predictions began to fail, one after another, the research field of AI largely withered.

It seems that AI is always predicted, by most experts, to be something like 15-25 years away from the present. Current experts often peg 2030 as a rough date for achieving artificial intelligence: just less than 20 years.

Will we have AI anytime soon? Based on the history of the field, and of predictions about it, the short answer is: I'm afraid we can't do that.

More technically involved analyses of the difficulties involved with emulating human intelligence lead to the same conclusion.

Computers excel in logical situations. They can rapidly determine the best course of action within any system of strict rules by brute force calculation of millions of options. However, no matter how fast their hardware becomes, they have never been able to adapt to the complexity of human reality. Computers excel at chess and checkers, but they are completely incapable of playing the game of life. (The real one, that is.)

Microchips have become exponentially faster over the past half-century, and software programs have become better at codifying reality and interacting with it. However, the closer you come to reality, the harder the next step is. While AI has made dramatic progress, it will be a very, very long time before we have conscious computer companions (or, perhaps, living in fear of our machine overlords).

(Image via Wikimedia Commons)

Sponsored Links

Tom Hartsfield
Author Archive