Will Machines Ever Truly Think? Richard Feynman Contemplates the Future of Artificial Intelligence (1985)

Though its answer has grown more com­pli­cat­ed in recent years, the ques­tion of whether com­put­ers will ever tru­ly think has been around for quite some time. Richard Feyn­man was being asked about it 40 years ago, as evi­denced by the lec­ture clip above. As his fans would expect, he approach­es the mat­ter of arti­fi­cial intel­li­gence with his char­ac­ter­is­tic inci­sive­ness and humor — as well as his ten­den­cy to re-frame the con­ver­sa­tion in his own terms. If the ques­tion is whether machines will ever think like human beings, he says no; if the ques­tion is whether machines will ever be more intel­li­gent than human beings, well, that depends on how you define intel­li­gence.

Even today, it remains quite a tall order for any machine to meet our con­stant demands, as Feyn­man artic­u­lates, for bet­ter-than-human mas­tery of every con­ceiv­able task. And even when their skills do beat mankind’s — as in, say, the field of arith­metic, which com­put­ers dom­i­nate by their very nature — they don’t use their cal­cu­lat­ing appa­ra­tus in the same way as human beings use their brains.

Per­haps, in the­o­ry, you could design a com­put­er to add, sub­tract, mul­ti­ply, and divide in approx­i­mate­ly the same slow, error-prone fash­ion we tend to do, but why would you want to? Bet­ter to con­cen­trate on what humans can do bet­ter than machines, such as the kind of pat­tern recog­ni­tion required to rec­og­nize a sin­gle human face in dif­fer­ent pho­tographs. Or that was, at any rate, some­thing humans could do bet­ter than machines.

The tables have turned, thanks to the machine learn­ing tech­nolo­gies that have late­ly emerged; we’re sure­ly not far from the abil­i­ty to pull up a por­trait, and along with it every oth­er pic­ture of the same per­son ever uploaded to the inter­net. The ques­tion of whether com­put­ers can dis­cov­er new ideas and rela­tion­ships by them­selves sends Feyn­man into a dis­qui­si­tion on the very nature of com­put­ers, how they do what they do, and how their high-pow­ered inhu­man ways, when applied to real­i­ty-based prob­lems, can lead to solu­tions as bizarre as they are effec­tive. “I think that we are get­ting close to intel­li­gent machines,” he says, “but they’re show­ing the nec­es­sary weak­ness­es of intel­li­gence.” Arthur C. Clarke said that any suf­fi­cient­ly advanced tech­nol­o­gy is indis­tin­guish­able from mag­ic, and per­haps any suf­fi­cient­ly smart machine looks a bit stu­pid.

Relat­ed con­tent:

Sci-Fi Writer Arthur C. Clarke Pre­dict­ed the Rise of Arti­fi­cial Intel­li­gence & the Exis­ten­tial Ques­tions We Would Need to Answer (1978)

The Life & Work of Richard Feyn­man Explored in a Three-Part Freako­nom­ics Radio Minis­eries

Isaac Asi­mov Describes How Arti­fi­cial Intel­li­gence Will Lib­er­ate Humans & Their Cre­ativ­i­ty: Watch His Last Major Inter­view (1992)

Richard Feyn­man Enthu­si­as­ti­cal­ly Explains How to Think Like a Physi­cist in His Series Fun to Imag­ine (1983)

Stephen Fry Explains Why Arti­fi­cial Intel­li­gence Has a “70% Risk of Killing Us All”

Richard Feyn­man Cre­ates a Sim­ple Method for Telling Sci­ence From Pseu­do­science (1966)

Based in Seoul, Col­in Marshall writes and broad­casts on cities, lan­guage, and cul­ture. His projects include the Sub­stack newslet­ter Books on Cities and the book The State­less City: a Walk through 21st-Cen­tu­ry Los Ange­les. Fol­low him on the social net­work for­mer­ly known as Twit­ter at @colinmarshall.