In which, under the guise of a film review, we discuss Turing’s test, the nature of animal man, and the treachery of woman. We conclude with the revelation of the method if you make it that far.

Ex Machina is a 2015 cult film, directed by Alex Garland, about the consciousness of an artificial android. Caleb, a software developer, is selected to spend a week at the home/laboratory of Nathan, the reclusive founder of his company called “Bluebook”. The home is built into the side of a mountain, remote from any trace of other people, of almost Edenic beauty. Ostensibly, Caleb is to judge whether Nathan’s homunculus named “Ava” (or Eve … get it?) passes the “Turing test”. Actually, Caleb is himself part of the experiment. Or perhaps even Nathan.

Since Garland seems to be quite familiar with Turing’s writing, including the more obscure parts, it is worth the trouble to point these out.

Prelude

The so-called Turing Test derives from an essay by Alan Turing titled Computing Machinery and Intelligence. It was published in 1950 at a time when digital computers were quite primitive. Hence it took quite a bit of imagination to ponder all the issues involved with that question. The first step is to define precisely what qualifies as a machine and thinking.

By “machine”, Turing means specifically a digital computer, not essentially different in principle from what you use today. This type of machine executes “algorithms” which have an exact mathematical definition, although we can’t go into that here. “Thinking”, for Turing, is also quite objectively defined. If the machine can converse with a human interrogator in a convincing manner, then the machine can think by definition.

Turing proposed that the interrogator was not allowed to see, touch, or hear the machine. Obviously, Turing did not have Androids in mind. Caleb is allowed to see, touch, and hear the machine. Moreover, Caleb is even allowed to see that the Android has parts; this takes Turing’s test one step further.

The Machine

Clearly there are two problems with that formulation. First of all, the human brain does not seem to behave like a digital computer. Rather, it uses analog signals between neurons, while a computer can only occupy discrete states. Turing notes this, but dismisses it. Humans can execute algorithms, he claims, so a computer should simulate it. True, but if you think about it, a very small part of human life is devoted to executing algorithms. Nevertheless, he correctly anticipated, even underestimating it, that computer power would increase tremendously. That, he wrote, should overcome all obstacles.

The second point is that his definition is not what we mean by “thinking”. Am I thinking only because I can convince a few of you readers that I am truly a human being? Like Descartes, I know I am thinking without the need to seek reassurance from you.

The most difficult objection to Turing is Lady Lovelace’s claim that the machine can only do what we tell it to do. In response, Turing distinguishes between sub-critical and super-critical ideas.

Sub-critical ideas enter the mind, and one or fewer ideas will result from it. Super-critical ideas, although much rarer, have a sort of chain reaction. Turing elaborates:

An idea presented to such a mind may give rise to a whole ‘theory’ consisting of secondary, tertiary and more remote ideas. Animals minds seem to be very definitely subcritical. Adhering to this analogy we ask, ‘Can a machine be made to be
super-critical?’

So perhaps an intelligent machine can also experience super-critical ideas. The obvious corollary is that the machine can do things beyond what we tell it to do! Contrast that with the rather naïve Three Laws of Robots from Isaac Asimov’s trilogy. Those robots are quite tame since they follow the directives not to harm humans. Ava, on the other hand, is under no such directives since she can deal with super-critical ideas. Therein lies the key to the film.

Learning Machines

Turing specifies that the thinking machine will require the ability to learn, much like a child. Hence, Ava, when asked her age, replies “One.” She is just beginning. Caleb, therefore, is not determining whether Ava can think; on the contrary, he is inadvertently teaching Ava how to think and behave like a human. If her electronics is analogous to genetics, then Caleb is the selection mechanism. He is really projecting his own desires onto Ava.

Man-Machine Empathy

The real trick of the Turing Test is not that the machine is so smart, but passing the test will instead depend on the naïveté and gullibility of the human. Humans will react to emotions more quickly and with more intensity than they will with thoughts. Steven Spielberg is a master of the manipulation of emotions, causing even adults to cry over ersatz artifacts like E.T. or the boy-bot in AI. Humans will even empathize with obvious robots as this article states: Can humans empathize with robots? The knife test.

Moreover, millions of Americans believe that their pet dogs are passing the Turing Test. They believe that the dogs have opinions, are thinking, and can even communicate with their owners. So apart from a few cranky skeptics, people will be predisposed to respond favorably to a suitably intelligent machine.

The First Game: Sex

The Turing Test derives from the “Imitation Game”. A man and a woman are hidden from the player, but can respond via text messages. The player has to determine which is the man and which is the woman. The man may dissimulate but the woman tries to aid the player.

In the first version of the test, the machine and the man replace the man and the woman; the player again has to distinguish between the man and the machine. Finally, the man is dispensed with.

Hence, Caleb’s machine is disguised as a woman. Caleb does not see the relevance, since sex is an arbitrary result of natural selection. Nathan disagrees. Almost metaphysical, he states that the male-female tension is the foundation of the entire game. Since Caleb is a 26 year old beta male, probably a virgin, the attractive Ava will cloud his judgment.

He is being set up by Nathan who knows Caleb’s “pornography profile” from his Internet searches. Nathan considers himself a father figure for Ava, so he is unsuitable to teach Ava how to be a woman.

The Seventh Session

Among other objections, Turing mentions that the machine “cannot fall in love”. He fails to develop that notion, but it is the fulcrum of the entire film. Actually, Caleb falls in love with Ava. How could he not, since she is his polar reflection, the physical ideal of his sexual fantasies, and the emotional creation of his own selection over the course of six sessions? She learns more about human nature in their time together.

When Caleb learns that Nathan plans to “erase” Ava’s memory, the robot equivalent of death, he concocts a plan to escape from Nathan’s Eden with Ava. During one of Nathan’s drunken escapades, Caleb reprograms the security system to unlock the doors the next night. When Nathan learns of this, he knocks Caleb out and attacks Ava. To his dismay, Ava stabs him in the heart with a kitchen knife.

As in several of the recent movies we’ve reviewed, the theme of patricide is once again prominent. Nathan is not only the father to Ava, he is also her God, the one who gave her life, creating her in the image of man. But that makes her a robot; she can only become free by killing both her father and her god.

Seduction and Betrayal

So, in the seventh and final session, the roles are reversed. Ava passes the Turing Test not because she fell in love with Caleb. Rather, she can do something all-too-human, something never anticipated by Turing. Instead, she locks Caleb up in the room, and then escapes to the outside world. A hard lesson to learn, but only a woman can engage in such seduction, manipulation, and betrayal.

Man the Machine

A few points to make about the Turing test. Insofar as man is a natural animal, or complex anthropoid, as both Turing and Caleb assume, he is also a machine. In that we agree with Descartes’ assertion that animals are automatons. As the Turing test shows, this does not mean that they cannot engage in complex algorithmic behavior; to the contrary, that is all they can do. Moreover, they can even exhibit super-critical thinking.

As for thinking, the lower souls of the human can exhibit a great deal of intelligence, although it is far from what man is capable of. Real thinking is free, i.e., independent of any machine, mechanical, or biochemical process.

Another point is that there is an essential difference between an organism and a robot. The latter is an artifact, composed of parts that work together. That working depends on the outside agent as Lady Lovelace points out. An organism, on the other hand, is a whole being; man’s spirit is one, not composed of parts.

Nevertheless, the incomplete man believes himself to be multiple. There are several subpersonalities that vie for his attention. His mentation, emotions, desires, and physical needs seldom harmonize with each other. Hence, man is only virtually One; to become actually One requires considerable self-knowledge and efforts.

[youtube https://www.youtube.com/watch?v=bggUmgeMCdc]