Lately, when I look out at the students in my college philosophy classes, I wonder whether they will end up using their minds to guide our future, or to build machines that will replace their minds and do that work for them. Certainly, they seem to use machines more and more to substitute for, or supplement, the functions of their minds. Daydreaming replaced by web surfing, flirting carried out by text messages, writing by cutting and pasting, making “friends?? on social networking sites—just to name a few. I can’t help wondering whether this is helping, or hurting, the development of their young minds.
As Artificial Intelligence makes new strides every day—creating machines that can play chess and Jeopardy, diagnose long-distance patients, as well as create ai machine learning that’s capable of learning on the go, and so much more—it is worth recognizing that the quest to build a true thinking machine is an old one, as are the worries that such a quest inevitably incites.
In 1770, a Hungarian inventor named Wolfgang von Kempelen brought a remarkable creation to the court of the Empress Maria Theresa—a life-sized mechanical automaton which, Kempelen said, could play chess. Known as “the Turk,?? the machine was demonstrated to acclaim throughout Europe and the Americas for nearly 84 years, in the process playing chess against such luminaries as Napoleon Bonaparte and Benjamin Franklin and causing audiences to marvel at the specter of a machine that could think as well as the greatest chess masters of the day.
But not everyone marveled—some dreaded the possibility of a thinking machine. A book published in 1784 related that “[o]ne old lady… went and hid herself in a window seat, as distant as she could from the evil spirit, which she firmly believed possessed the machine.??
Yet the age of the thinking machine had not in fact yet come: in 1821 the Turk was revealed to be a grand hoax. A human operator was hidden inside the mechanism, controlling the Turk’s chess pieces by the use of magnets under the chess board. Ironically, the true start of the age of artificial intelligence came at the same time that the hoax was unveiled.
That same year, the brilliant mathematician Charles Babbage began thinking of inventing a calculating machine. His Difference Engine—never fully built in his own time—marked the beginning of artificial intelligence.
Made of brass and steel, with figure wheels representing the digits, Babbage’s invention could calculate any polynomial function set in at the start, and did so with no further intervention from the human operator save the turning of a crank handle. The Difference Engine would have been the first machine to supplant not physical labor—like other recent inventions, the steam engine, the mechanize loom, the locomotive, or the cotton “gin?? (short for “engine??)—but mental labor. Had the machine been built, it would have rendered obsolete the hoards of men and women who calculated tables of mathematical numbers, such as logarithm tables, astronomical tables, and actuarial tables. These calculators were called “computers,?? giving rise to the name of the machines that did, eventually, supplant the human table calculators.
That Babbage’s invention marked the start of a new era was recognized immediately, even before he began to construct the engine. The Astronomical Society, which awarded Babbage a gold medal, described his Difference Engine as a machine that “substitutes mechanical performance for an intellectual process.?? A friend of Babbage’s more vividly remarked that in his machine “the marvelous pulp and fibre of a brain had been substituted by brass and iron, [Babbage] had taught wheelwork to think.??
For Babbage and his friends, the idea that a machine could do the job of human intelligence was empowering. Mathematicians could be freed up for more important work while ensuring a level of accuracy that no human computer could hope to attain—the tables that were calculated by hand were rife with error. But others of his time were not so sanguine about the possibility of artificial intelligence. To them, Babbage’s engine raised troubling questions about the value of human minds, just as the mechanized looms and other inventions had raised questions about the value of human labor, questions that would soon be taken up by a young Karl Marx. If machines could think, were human minds no better than machines? And if a machine could be made to perform the mental functions of mathematical reasoning, were other mental functions far behind? Language, emotions, creativity—would these come next?
The very idea of a Difference Engine provoked much anxiety in the 19th century, not only among the general public but also those engaged in scientific research, and this is one reason Babbage’s machine was never built in his own day.
Just as in Babbage’s day, the specter of a thinking machine worries us. “Long live humanity?? one headline declared defensively, when the physicist and New Jersey congressman Rush Holt won one round against the Jeopardy-playing computer Watson. But Babbage’s failure to build his thinking machines did not stop the quest to create one. Before we get there, we should spend some time thinking about the consequences, both for society and for individual human minds. An od advertisement used to say, “A mind is a terrible thing to waste.?? Might it not, also, be a terrible thing to replace?