Alan Turing & the Imitation Game

alanturingHobyahs2smallThe Imitation Game is released in local theaters on Friday, so it seems a good time to discuss the philosophy of mind of the film’s protagonist, Alan Turing.

The imi­ta­tion game, or the Tur­ing test, is a test in which a com­put­er is pit­ted against a human in the task of con­vinc­ing anoth­er human that it, the com­put­er, is also a human.  The test is often con­strued as a means of gath­er­ing evi­dence that the com­put­er is or is not intel­li­gent, but this is a mis­rep­re­sen­ta­tion of Turing’s actu­al posi­tion.  Tur­ing is best under­stood as hav­ing main­tained that being treat­ed as intel­li­gent, as hav­ing a mind, is con­sti­tu­tive of hav­ing a mind.  Pass­ing the Tur­ing test was not intend­ed to mere­ly indi­cate that the com­put­er has a mind but rather to make it true that the com­put­er has a mind.

Turing’s test is an attempt to solve a cen­turies old philo­soph­i­cal puz­zle: the prob­lem of oth­er minds.  I know that I have a mind because I have priv­i­leged, first-per­son access to it.  But I can­not know that you have a mind in the same way.  I could see that there are cer­tain things that are true of me, e.g., I have a brain, I react in such-and-such a way to such-and-such a stim­u­lus, etc., and infer that if those are true of oth­ers, then they too have a mind.  But to make that infer­ence I must know that oth­ers are sim­i­lar to me in the rel­e­vant respects, that what is asso­ci­at­ed with hav­ing a mind in my case is asso­ci­at­ed with hav­ing a mind gen­er­al­ly.  I can­not know that on the basis of one case alone, i.e., my case, and I do not have access to any oth­er cas­es.  It is puz­zling, then, how we can come to know that oth­er peo­ple have minds.

Dur­ing much of the peri­od of Turing’s life, the 1920’s and 30’s in Eng­land and the 40’s and 50’s in the U.S., the philo­soph­i­cal move­ment of log­i­cal pos­i­tivism (or log­i­cal empiri­cism) was dom­i­nant.  Log­i­cal pos­i­tivism is real­ly a fam­i­ly of posi­tions, but they have in com­mon the goal of pro­vid­ing a strict stan­dard for the mean­ing­ful­ness of claims, par­tic­u­lar­ly philo­soph­i­cal claims.  Famous­ly, many log­i­cal pos­i­tivists reject­ed claims like “God exists”, “God does not exist”, “It is wrong to lie”, and “It is not wrong to lie” as nei­ther true nor false but rather mean­ing­less because there is no pos­si­ble way to ver­i­fy or fal­si­fy such claims.

Since the claim that some­one has a mind is nei­ther prov­able nor fal­si­fi­able if we under­stand mind to be some unob­serv­able enti­ty, the British philoso­pher Gilbert Ryle argued that hav­ing a mind just is behav­ing as though you have a mind.  This would make the claim that one has a mind empir­i­cal­ly ver­i­fi­able since behav­ior is observ­able.  Bonus Points: Iden­ti­fy the most sig­nif­i­cant prob­lem with this solu­tion to the prob­lem of oth­er minds.

In this philo­soph­i­cal atmos­phere, it makes per­fect sense that Tur­ing would take pass­ing the imi­ta­tion game to be one way to have a mind.  “Imi­ta­tion game” is, how­ev­er, a mis­nomer since the com­put­er, if it pass­es the test, is not mere­ly imi­tat­ing some­thing with a mind; it is some­thing with a mind.  Tur­ing wrote:

The orig­i­nal ques­tion, “Can machines think?” I believe to be too mean­ing­less to deserve dis­cus­sion. Nev­er­the­less I believe that at the end of the cen­tu­ry the use of words and gen­er­al edu­cat­ed opin­ion will have altered so much that one will be able to speak of machines think­ing with­out expect­ing to be con­tra­dict­ed.

Remem­ber this the next time you’re wait­ing for your com­put­er to com­plete a func­tion and you say to your friend, “It’s think­ing.”

Tur­ing, Alan. 1950. “Com­put­ing Machin­ery and Intel­li­gence.” Mind 59(236): 433–60.

Tur­ing, Com­put­ing Machin­ery and Intel­li­gence

Bookmark the permalink.

Leave a Reply