Lecture 2: The Imitation Game: Understanding Turing’s Test for Intelligence

Added: 2026-05-11

[00:00:17]Hello everyone, my name is Dr. Amar Singh. I'm assistant professor in English at MMB in Banaras Hindu University.

[00:00:28]I welcome you all in the second lecture of this course understanding AI through literature and popular culture.

[00:00:37]Now in this particular lecture we'll cover certain objectives.

[00:00:42]We'll talk about Alen Turing's imitation game and explain how the test is played.

[00:00:50]We will reflect on whether imitation in machines can or cannot lead to genuine self-awareness.

[00:00:59]And also we'll recognize how the imitation game reveals and reinforces hierarchal divisions a dynamic still visible in present context.

[00:01:14]So you guys must have heard about this particular term the during test in some way or in some form in some film in some novel this particular term must have come across. Now what this during test means? So during test has become a benchmark where a machine if it can act as a human if it can imitate human in such a manner that even the human subjects are deceived whether they are interacting with a human or a non-human entity then that machine has actually passed the Turing test.

[00:01:53]So before John McCarthy and Marvin Minsky when they coined this particular term artificial intelligence in 1956 there already existed the concept of intelligence in machine which was known as spoiler alert machine intelligence.

[00:02:12]So in 1950 Alan Turing he wrote a paper called computing machinery and intelligence.

[00:02:22]He begins this paper by asking a question. Can a machine think?

[00:02:29]Now there are two things in focus here.

[00:02:33]There is machine and there is think. So thinking is a thing which is basically done by humans.

[00:02:43]So how can a machine think? Machine is usually fed with data and it interprets those data. It goes through symbolic manipulations.

[00:02:54]So how can a machine think?

[00:02:59]So Turing comes with a thought experiment where he says that in future this whole point whether a machine can imitate human will become moot. But before that u he comes with this proposition of imitation game.

[00:03:17]As in this particular game, the machine will try to imitate humans. But machine will come later on. So we'll see what happens here. So there are three players in this particular game. There is A, B and C. So A is man, B is woman and C is basically an interrogator who could be of either sex.

[00:03:43]Now all these participants they are set apart they are sent in different rooms. So A is another room, B is in another room and the interrogator stays in another room.

[00:03:57]Now the interrogator will ask certain questions from A and B. The job of the interrogator is to decipher whether A is man or A is woman and vice versa whether B is man or B is woman.

[00:04:15]What A will do is that A will try to A will try to trick the interrogator.

[00:04:22]A will uh deceive the introgator by saying that A is not man, A is a woman.

[00:04:33]On the other side, B who is basically woman, she has to stay truthful. She has to answer correctly.

[00:04:42]So if a question is put by the introter so to remove any kind of voice modulation voice answer all the answers are to be typewritten.

[00:04:58]So if a if the interrogator asks define yourself or categorize yourself or tell me some attribute of yours. So A will try to present himself as a woman. A will say that uh he likes pink.

[00:05:15]He may say that his hair is long.

[00:05:20]Whereas on the other side, woman who is B will she could simply say that A is wrong.

[00:05:29]A is man. She is the woman. She likes pink. She may like blue. Whatever the case is, but she will answer everything truthfully.

[00:05:40]So Turing says what if we remove the man from this picture who is basically a and we place a machine there.

[00:05:54]Now this whole test will be done between machine woman and the tester.

[00:06:05]Thereby A is machine, B is woman and C is tester.

[00:06:11]So we'll see how like how things proceed. There's an excerpt from the paper that I have taken here where the introator is asking questions.

[00:06:20]We'll see how kind what kind of questions are being asked.

[00:06:24]So let's first go through uh some of the points in the slide here.

[00:06:30]When the question is posed in this manner which is like can the machine think it places a great deal of emphasis on the definition of the terms machine and think.

[00:06:40]Turing recommends replacing this question with another that is closely related which is can machine imitate.

[00:06:51]According to him this new problem could be compared to a game which he refers to as an imitation game.

[00:06:59]This game is played between three people.

[00:07:03]A man A, a woman B, and an intro C who may be of either sex.

[00:07:14]The interrogator stays in a room apart from the other two.

[00:07:19]The job of the interrogator is to determine which of the other two is the man and which is the woman.

[00:07:27]He knows them by the labels X and Y.

[00:07:31]In the end, he has to determine that either X is A and Y is B or vice versa.

[00:07:39]So the interrogator might put a question like will X please tell me the length of his or her hair?

[00:07:50]Now suppose X is actually A.

[00:07:54]Then A must answer. It is A's object in the game to try and cause C to make the wrong identification. I've highlighted this portion. What is his job? What is A's job here? That he has to cause C to make the wrong identification.

[00:08:12]His answer might therefore be my hair is shingled and the longest strands are about 9 in long.

[00:08:22]So we can see clearly here that he's trying to deceive the introgator. He's basically putting inferring that his hairs are long and this might mean that A is woman.

[00:08:36]In order that tones of voice may not help the introgator, the answers should be written or better still typewritten.

[00:08:46]The ideal arrangement is to have a teleprinter communication between the two rooms.

[00:08:52]Alternatively, the question and answers can be repeated by an intermediary.

[00:08:59]The object of the game for the third player, which is B, is to help the interrogator. So the woman has to help the interrogator.

[00:09:08]The best strategy for her is probably to give truthful answers.

[00:09:15]She can add such things as I am the woman. Don't listen to him to her answers. But it will avail nothing as the man can make similar remarks.

[00:09:28]We now ask the question what will happen when a machine takes the part of a in this game. So the machine replaces the man.

[00:09:38]Will the interrogator decide wrongly as often when the game is played like this as it does when the game is played between a man and a woman.

[00:09:49]So these questions replace our original can machines think. So what could be the new kind of question? Again spoiler alert. Can machines imitate?

[00:10:00]Now moving forward, Turing says that in the future as this paper was basically written in 1950. So moving forward coming in 21st century in the contemporary era that we are living in we are surrounded with technology and Turing says that as technology gets advanced uh as it becomes better faster powerful the very question whether machine can imitate humans will become moot and we can sense something like this happening we are surrounded with AI technology we use chat GPT people do share their inhibitions their their fear uh the happiness with charge. They treat it as if charge is uh their psychiatrist.

[00:10:45]Students treat it as their teacher. Uh patients they put uh their medications and ask for uh deciphering what kind of salts are used whether it will have a good effect or bad effect on their body.

[00:10:59]So the thing is that people are actually not treating machine uh in terms of whether it is imitating humans or not.

[00:11:07]So something which Turing predicted so many years back is now coming to fruition.

[00:11:15]So Turing asserts in this paper that in the future when technology advances to a certain level and beyond the question of whether machines can think will become a mood point.

[00:11:28]So I I'll quote him.

[00:11:30]I believe that in about 50 years time it will be possible to program computers to make them play the imitation game so well that an average intro will not have more than 70% chance of making the right identification after 5 minutes of questioning. The original question, can machines think? I believe to be too meaningless to deserve discussion.

[00:11:58]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. So now we are actually doing that. We are not talking about whether machine is imitating humans. We are actually talking about machines in terms of like how well they are thinking.

[00:12:24]So moving forward, Turing also brings in this paper the very question uh that becomes the burning point when it comes to uh artificial intelligence whether consciousness may play some role moving forward.

[00:12:43]So he uses professor Jefferson's uh quote and I'll quote quote him here now. He says not until a machine can write a sonnet or compose a concerto because of thoughts and emotions felt and not by chance fall of symbols could we agree that machine equals brain.

[00:13:10]That is not only write it but know that it had written it.

[00:13:16]No mechanism could feel pleasure at its success, grief when its wolves fuse, be warmed by flattery, be made miserable by its mistakes, be charmed by sex, be angry or depressed when it cannot get what it wants.

[00:13:35]So we'll just go through what this particular quote means.

[00:13:40]The mere fact that a machine can produce words, music or art does not equate it to the human brain.

[00:13:48]If it is to be like the brain truly, it must create based on real feelings and thoughts and not simply on a combination of symbols.

[00:13:59]Furthermore, the machine must also comprehend that it created the output, not merely producing it mechanically.

[00:14:09]Unlike humans, machines are not capable of feelings things like joy when they succeed, sadness when they break down, pride when praised, shame or misery when they fall, attraction or or love, anger or frustration when they don't get what they want. So in short, machines can simulate actions but cannot have self-awareness. So what he's trying to basically say here that people artist poets uh novelist when they write they do it in in the fit of emotions.

[00:14:46]So say I'll use an example here when Shelley writes um I fall upon the thorns of life I bleed a heavy weight of ours have have chained and bowed or when Arnold says when he's on a honeymoon with with his beloved wife and he's looking at do beach and suddenly the scene uh brings so much of emotion in him such profound emotions that he confesses his love in the manner all love let us be true to one another in a world which seems to lie before us like a land of dreams. So when people feel certain sense of emotion in them they produce art.

[00:15:30]So if machine is doing something like this can we actually say that it is producing art that it is actually feeling what it is writing.

[00:15:41]So the very question that it is not enough that you are doing the act you should also be aware of that you are doing the act you know so there's a certain detachment you know of the very action that you are doing does machine know that can machine understand something like this not only participate but also be aware of what John Keats says in terms of negative capability living within the uncertainties living within the paradox boxes can a machine do something like this. So Turing then again contradicts he says that why should we confine ourselves into thinking that a machine which is an electronic species should think in terms like humans who are biological beings.

[00:16:30]According to Turing the particular argument appears to challenge the validity of his test.

[00:16:38]If one must go on this route, then the only way to understand what machines think is by being the machines and feeling oneself think.

[00:16:50]This would create a solstic point of view in which for example to know what a man thinks would be to be that particular person. This exacerbates the difficulty of communicating ideas.

[00:17:04]As a result, A would think that B does not think while B would think that B does think but A does not.

[00:17:12]So rather than arguing about this, one can assume that everyone thinks. So again, we already dealt in the previous lecture where we saw how west is actually like it uh deals with the notion of hierarchy when it come when it comes to intelligence.

[00:17:32]Especially white men when they see uh humans of uh uh other nations, non-white people, uh people of color, uh they assume that their intelligence is on the lower part. That's how West has always treated.

[00:17:50]So they might have been thinking or they I don't know whether they are still thinking in those terms or not but the history has shown that people have treated other people in terms of that they are the ones who think but the other party does not. A man thinks but a woman does not. Woman thinks that she feels and she thinks a lot but man does not. So Turing comes and says that why to go into this roundabout thing and we should come together and can agree that everyone thinks even machine thinks if machine can possess intelligence it might be interpreting it whatever data is being given to it it might be interpreting it in its own terms so machine thinks.

[00:18:37]So again I have a few questions for all you people and I am looking forward to your discussion in the discussion forum uh for some amusing answers for some amusing interpretations.

[00:18:48]So the first question is how far is Ellen Turing's insight being realized today in terms of machines mimicking humans more accurately. So when you are chatting with Chad Jupiti, do you think that Chad Jupit is actually imitating like humans or maybe even better?

[00:19:08]The second Turing designed his thought experiment in the form of an imitation game in which it was incumbent upon the woman to prove her identity whilst competing with the machine. So we saw that the man in that tripotite he detached himself from the action and all that was left was B which is like woman and C the inrogator who could be of either sex. So man detached himself but still is a main component of the whole play but is watching everything objectively.

[00:19:48]So the woman here is competing with the machine. Woman has to prove herself her identity while competing with the machine that the machine is not the woman but woman is the woman.

[00:20:01]Would it be possible to view the imitation game as a prediction by Turing in which he predicted that the burden would always fall on the person struggling within the hierarchy.

[00:20:13]Now with these questions I would take leave. Thank you all of you.