Fantasy Science Pt. 2: The Turing Test & EX MACHINA
Networks. Artificial Intelligence. Robotics. Have you heard terms like these flying around the science fiction sections of the film world? Have you ever wondered just how accurately these films portray real science? Well, my friends, today is your lucky day: this column, Fantasy Science & Coffee, aims to bridge the gap between science and science fiction in films and occasionally popular culture. My hope is to explain things in a fun way – like we’re chatting over coffee.
You may be thinking: who is this person, why does she think she can explain science, and why the heck would I want to have coffee with her? Well, I’m Radha, a researcher in India, currently pursuing a PhD in theoretical quantum physics. I quite like hot beverages. I’ll also pay.
In this second part of the series, I’m going to look at how Ex Machina (2015) portrays something called the Turing Test.
Let’s begin.
If you are even remotely familiar with technology, science fiction, or the world of computers, you’ve probably heard the phrase ‘Turing Test’, even if you don’t quite know what it means. It is essentially a test to see whether an Artificial Intelligence can mimic human behaviour flawlessly, essentially achieving ‘sentience’.
The basis of the 2015 film Ex Machina is the Turing test. In it, a genius multi-billionaire programmer, Nathan, develops a fully functional Artificial Intelligence called, Ava, in humanoid form. One of his employees, Caleb, is invited to his estate to test just how lifelike Ava is. The test that we see on screen is a more sophisticated version of the original Turing test, and a more appropriate form of testing given the extent of technology there is in the story. To illustrate just what I mean, let’s first understand Turing’s initial idea.
The Original Turing Test
In 1950, the renowned mathematician, 2014 biographical film about him with the same name.)
In it, he strove to answer the question: “Can machines think?” To answer this, he proposed a test consisting of three parties: A) a human examiner, B) a human test subject, and C) a machine test subject.
B and C are placed in such a manner that the examiner does not know which is machine and which is human. If C can answer questions that the examiner poses to them in such a manner that the examiner cannot tell the difference between human and machine, then the machine has successfully demonstrated thought.
It is a straightforward test, and the crux lies in the machine’s ability to mimic human speech. During Turing’s time, a machine was far from being able to attain such mimicry. Yet, today, amazing strides have been made on this front, and Turing’s original criteria seem redundant, or at the very least, too simplistic.
With the advent of chatbots, or bots that mimic human speech that humans can interact with, it’s clear that moving beyond the original test is needed. These days, chatbots are not only able to imitate human speech, they are used everywhere. Those of you who are on Tinder may have even found yourself chatting with one!
This issue has been addressed in a cute, short video by TED-Ed:
While plenty of people have proposed alternate versions of the test, we stick to the one in Ex Machina for now, because it’s pretty darn cool. Admittedly not as cool as the XKCD Turing Test, though.
The Turing Test in Ex Machina
Let’s take a peek at an interaction towards the beginning of the film between Nathan and Caleb. Here’s a snippet from Alex Garland’s screenplay:
This exchange takes place towards the beginning of the film, and Caleb is a little stuck on the nature of the test he should perform. You see, unlike the traditional Turing test, Caleb knows exactly what Ava is. It appears at first glance that everything is out in the open and there is no level of concealment between the examiner and examinee.
However, the catch is that there actually is a hidden screen between the two, which is revealed at the end of the film. Caleb is actually a part of the test, and is not the examiner. Nathan programmed Ava with a full spectrum of human emotions and sexuality, and led her to believe that she is an abused captive with only Caleb as a means of escape. Let’s look at this exchange towards the end:
Fast forward a couple of minutes later, and we see:
The beauty of the test result in this film is that Ava behaved exactly as any human prisoner would have. Using tools of manipulation, she led Caleb to feel sorry for her and develop feelings for her strong enough for him to trick Nathan and arrange her escape. Little did Caleb know that he was signing both his and Nathan’s death warrants with that empathy.
Mimicking human speech is old news. If we want to show just how lifelike a machine can be, a sophisticated Turing Test such as the one in Ex Machina is required, on par with the level of technology today. While I find the thought academically intriguing, seeing Ava coldheartedly leave Caleb to die was, frankly, frightening. Did it scare you too?
More to Explore
Articles
ACM: Moving Beyond the Turing Test with the Allen AI Science Challenge (2017)
Huffington Post: The Turing test is tired. It’s time for AI to move on (2017)
ZDNet: MIT’s artificial intelligence passes key Turing test (2016)
IEEE Spectrum: Artificial-Intelligence Experts to Explore Turing Test Triathlon (2015)
BBC News: Computer AI passes Turing test in ‘world first’ (2014)
Live Science: New Artificial Intelligence Challenge Could Be the Next Turing Test (2014)
Papers
Alan Turing’s original paper: Computing Machinery and Intelligence [PDF] (1950)
Internet Resources
The Alan Turing Internet Scrapbook
For a bit of fun, try having a chat with CleverBot: http://www.cleverbot.com/
Does content like this matter to you?
Become a Member and support film journalism. Unlock access to all of Film Inquiry`s great articles. Join a community of like-minded readers who are passionate about cinema - get access to our private members Network, give back to independent filmmakers, and more.
