We take back the awful things we said about Skynet
‘Do you believe you have a soul?” The question is delivered to what is normally considered a soulless, non-sentient object in the trailer for a new film. Is it Ultron, the megalomaniacal artificial intelligence in the new Avengers movie? Ava, the robot from Ex Machina? Arnold Schwarzenegger in a rare, quiet moment from Terminator: Genisys?
Nope. It ’s Ted, the foul-mouthed Teddy Ruxpin lookalike voiced by Seth MacFarlane in Ted 2. In the sequel to the 2012 comedy, a teddy bear who came to life when his owner was a child has grown up and wants to start a family. Lacking DNA (and possibly also a— well, let’s not go there) he seeks a sperm donor in Mark Wahlberg’s character.
But then comes this shocker, delivered by Ted’s (human) wife: “If we want to have a baby, you’re going to prove you’re a person in a court of law.”
Clearly, this represent a zeitgeist moment in filmmaking. Avengers: Age of Ultron, Ex Machina and Terminator: Genisys all grapple to varying degrees with the notion of machine intelli-gence, and what it means to the future of the old-fashioned, biological kind. So do The Machine and Autómata, two recent science-fiction films that didn’t get theatrically released in Canada. But these are merely the tip of the circuit board.
The recent release Chappie, from writer/director Neill Blomkamp, imagines a scientist who imbues one of his robot police officers with selfawareness. Her, from 2013, featured a writer who falls for his computer operating system. Disney’s Big Hero 6 featured a robotic health-care assistant who seems to have self-awareness and perhaps even that ineffable human construct, a soul.
Of course, “seems to” and “perhaps” are galaxy-spanning wiggle words. In 1950, computer pioneer Alan Turing imagined what he called “the imitation game.” It has since become known as the Turing test. Put simply, if you can have a conversation over email with a computer and be unable to conclude that it not a person, then the machine has won the game and passed the test. Whether it is also self-aware has occupied philosophers, semioticians and AI researchers in the decades since.
A Turing test is the central plot point in Ex Machina. Programmer Caleb is asked by his boss to converse face-to-face with Ava and decide whether she/it is self-aware. It’s a twist on the classic test, since the subject is not trying to deceive the tester that she’s human. Ava is a machine, though a sexy one. But is she aware?
One of the many problems with the Turing test is subjectivity and bias on the part of the tester. We humans anthropomorphize everything, from the weather to our pets and, increasingly, our technology. A Washington Post story from 2007 describes a U.S. Army colonel who stopped a test of a robot mine-clearer as the damaged bot dragged itself forward on its one remaining leg. The colonel deemed it “inhumane.”
More recently, MIT roboticist Kate Darling let participants at a conference play with an animatronic dinosaur named Pleo. After an hour of interacting with the cute robot, which responds to voice commands and touch, they were asked to destroy it. No one would. When she said she would destroy all the Pleos unless one was sacrificed, a volunteer executioner reluctantly stepped forward.
No one is suggesting that Pleo can feel, and yet it speaks to our inherent sense of what might be deemed machine fairness that we don’t want to harm something the looks like it can feel. And if we get tied into such ethical knots at the thought of dismembering a piece of plastic, what happens when Pleo evolves past the robot Pleistocene?
Films present a unique forum for exploring ideas about artificial intelligence, because they already play a kind of imitation game with audiences. No one believes that an image on celluloid has a soul, and yet we cry real tears when they suffer. We feel for them, though they feel nothing. Wh e n Leonard Nimoy died, eulogists spoke not only of the man but of the nonentity that was the character of Spock.
Conversely, films often fail to connect with audiences when they fail the cinematic Turing test. We speak of chemistry between characters, of motivation, believability and consistency. Would Indiana Jones ever join the Nazi cause? The correct answer is: Of course not, because he’s fictional. But the answer that springs to mind is: Of course not, because he’s Indiana Jones.
In The New York Times, Ex Machina writer/director Alex Garland suggested that the glut of AI films should, given Hollywood production cycles, stem from an event in that field some three years ago, yet he found nothing. I would argue that he needed to look further back, to 2011, when IBM’s Watson beat two human contestants at the game show Jeopardy!
At the time, third-place finisher Ken Jennings joked that he would “welcome our new computer overlords.” The refrain has since been taken up in a more serious vein by such thinkers as physicist Stephen Hawking and entrepreneur Elon Musk, who warn that AI has the potential to create a dark future if machines are not kept under strict control.
The problem with control is that it raises even more ethical quandaries. In Ex Machina, Isaac’s character describes himself as a father to his creation. Biological parents have been making intelligent, self-aware creations for millennia. We imprison them in our homes, regulate their lives and control their existence, and this is deemed good and natural — up to a point. Continued too long, however, it becomes slavery.
Humans may one day need to decide when their technological offspring reach the age of reason, consent and freedom. It will be a test, though hardly a game. Fortunately, if the recent history of film is any indication, we have already started to give it serious (or in the case of Ted 2, comical) thought.