I finally got around to seeing Ex Machina, a new Science Fiction movie in theaters. Ex Machina is an intriguing fictional inquiry into the philosophical question of whether machines could ever be truly intelligent. And if so, how could we tell? If you haven't seen the movie yet, then watch this trailer and be warned that this post will contain some spoilers, so read at your own risk.
Ex Machina qua Movie
The general background of the movie is as follows: a bright young professional named Caleb is invited by the founder and CEO of his company (Blue Book, a fictional search-engine which parallels real-life Google), Nathan, to his private home/research facility for a week. Once there, he learns that his employer might have created the first Artificially Intelligent machine. Nathan reveals to Caleb that he has been selected to test the new development, Ava, to determine whether "she" really is intelligent and self-aware.
Purely in terms of cinematic quality, this is a fantastic movie. Director/writer Alex Garland does a great job of unfolding this narrative by introducing necessary information in an intentional and artful way. Although the movie is almost entirely dialog with very little action, it kept me engaged for the entire two hours. One of the biggest strengths of the movie is the way with which the "science" is dealt. Because Science Fiction is... well... fiction, it is difficult to make it seem realistic. Any movie that tries to give too detailed an account of how the science in the movie works is taking a huge risk. The moment they slip up and make a mistake, any science-minded individual (a majority of the target audience) will likely be too distracted to enjoy the movie. Ex Machina leaves enough up to the imagination that it seems realistic and organic, despite dealing with futuristic themes. The acting is also very good, as nothing seems forced or out of place. I have one disclaimer though: there are a few scenes with nudity in them, so you may not want to watch it if you are sensitive to that.
Ex Machina qua Philosophical Commentary
One of the reasons that this movie appealed to me is that, in addition to the scientific themes, the movie grapples in an impressive manner with some real philosophical concepts. Indeed, the movie might be seen as a fictional inquiry into the field of Philosophy of Mind. Questions like "can machines ever truly think?", "what is a mind?", and "can mechanistic processes ever exhaustively explain what it means to experience the world?" are raised by this plot.
While I applaud the movie for bringing up these questions and making them accessible to those who may not have had any exposure to them, I think ultimately the movie leans toward some hasty answers. The philosophical theme is summed up early on in the film with the discussion of the Turing Test.
The Turing Test is a method, that was proposed by philosopher/mathematician Alan Turing (star subject of the recent film The Imitation Game), to determine if a machine has become intelligent. Imagine that you are communicating via a keyboard and mouse with two different "things." One of these is a normal human being, while the other is a machine with possible artificial intelligence. The machine passes the test-- meaning it can be said to have a "mind," and be self-aware-- if you cannot tell which is the human and which is the machine. This is the task that Caleb is asked to perform, with some differences.
My issue here is that the question has been framed wrong from the start. The movie (along with many real-life philosophers and scientists) would lead us to think that this is a way with which we can prove Strong AI. While the term "Weak AI" is generally used for something that can perform like it has a mind, "Strong AI" is reserved for the Real McCoy. So if we say that something has achieved Strong AI, then we are suggesting that it has a real mind, with real self-awareness, just like human beings.
The movie is set up from beginning as if Turing's philosophy was right-- namely that simply being able to indistinguishably act like a human proves that a machine has Strong AI. But this is really just Philosophical Behaviorism-- the idea that emotions, feelings, sensations, and experiences can be explained by referencing nothing but the behaviors which accompany them. So, for example, if someone reacts as if they are in pain, then they are in pain. If someone is uncontrollably crying, then they are sad. Emotions, thoughts, and intentions are reduced to the behaviors that they manifest. The problem is that this doesn't really show that someone is in pain, or sad. It is just a useful way for us to assume that they are those things without having access to their internal mental states. If it did, there wouldn't be such a thing as acting. That human beings have an entire entertainment industry based on acting implicitly shows that we know that how a person behaves is not necessarily indicative of their internal states.
So let's turn this back to Ex Machina. The movie wants us to buy into the fact that we should consider a machine to be conscious, thinking, and feeling, as long as it can behave as if it is conscious, thinking, and feeling. To its credit, the movie appeals to behaviors much more complex than simple crying. In the film we are to believe that Ava is conscious because she can tell a joke, appear to be falling in love with Caleb, cause Caleb to fall in love with her, come up with an intricate plan of escape which involves deception, etc. But the problem is that no matter how complex these criteria are, they still fail as tests of consciousness for the same reason that my crying fails as a test that I am sad. Of course I often do cry when I am sad, but I am more than capable of creating crocodile tears. And a complexly programmed machine could do the same.
So while the movie is overall excellent, the plot has some misguided philosophical underpinnings.
0 comments :
Post a Comment