By 2024, AI will make the news every day. We may be aware of the science, but how do we imagine AI and our relationship with it, now and in the future? Fortunately, film can give us some insights.
Probably the best-known AI in film is HAL 9000 from Stanley Kubrick’s 2001: A Space Odyssey (1968). HAL is an artificially intelligent computer housed aboard a spacecraft capable of interstellar travel. The film was released less than a year before man landed on the moon. And yet, even amid this optimism about a new era of space travel, HAL’s portrayal sounded a note of caution about artificial intelligence. His motivations are ambiguous and he proves capable of turning against his human crew.
This 1960s classic demonstrates the fear common throughout the history of AI cinema: that AIs cannot be trusted, that they will rebel against their human creators and try to overpower or overthrow us.
These fears are contextualized differently in different historical eras: in the 1950s they are associated with the Cold War, followed by the space race in the 1960s and 1970s. In the 1980s it was video gaming, and in the 1990s it was the Internet. Despite these various concerns, fear of AI remains remarkably consistent.
My latest research, which forms the backbone of my new book AI in the Movies, examines how ‘strong’ or ‘human’ AI is depicted in film. I examined more than fifty films to see how they shed light on human attitudes toward AI – how we interpret and understand it through characters and stories, and how attitudes have changed since AI’s inception.
Types of AIs
The idea of AI emerged in 1956 during a US summer research project workshop at Dartmouth College in Hanover, New Hampshire, where a group of academics met to brainstorm ideas around ‘thinking machines’.
A mathematician named John McCarthy coined the name “artificial intelligence” and as soon as the new scientific field had a name, filmmakers were already imagining human-like AI and what our relationship with it might be. The same year, an AI, Robby the Robot, appeared in the film Forbidden Planet, and the following year in 1957, he returned in the film The Invisible Boy to defeat another type of AI, this time an evil supercomputer.
The AI as evil computer reappeared in 1965 as Alpha 60, in the chilling dystopia of Jean-Luc Godard’s Alphaville, and then in 1968 with Kubrick’s memorable HAL in 2001: A Space Odyssey.
These early AI films set the template for what would follow. There were AIs with robot bodies and later robot bodies that looked human – the first of these appeared in Westworld in 1973, where a robot malfunction in a futuristic amusement park for adults causes chaos and terror. Then there were AIs that were digital, like the evil Joshua in the 1977 horror film Demon Seed, in which a woman is impregnated by a supercomputer.
In the 1980s, digital AIs began to become connected to network computing – where computers “spoke” to each other in an early incarnation of what would become the Internet – like the one Matthew Broderick’s high school student encountered in War Games (1983). ), who almost accidentally starts a nuclear conflict.
From the 1990s onwards, an AI could move between digital and material domains. In the Japanese animated film Ghost in the Shell (1995), the Puppet Master exists in the ebb and flow of the Internet, but can inhabit ‘shell’ bodies. Agent Smith in The Matrix Revolutions (2003) takes over a human body and materializes in the real world. In Her (2013), the AI operating system Samantha ultimately goes beyond matter, beyond the ‘things’ of human existence, and becomes a post-material being.
Mirrors, doubles and hybrids
In the early decades of AI cinema, AI characters mirrored human characters. In Collosus: The Forbin Project (1970), the AI supercomputer reflects and reinforces the inventor’s arrogant ambition. In Terminator 2: Judgment Day (1991), Sarah Connor has become like the Terminators of AI Skynet itself: her power is her armor and she hunts to kill.
By the 2000s, human-AI twins began to overlap and merge into each other. In Spielberg’s AI: Artificial Intelligence (2001), the AI ”son” David looks exactly like a real boy, while the real son Martin comes home from the hospital connected to tubes and wires that make him look like a cyborg.
In Ex Machina (2014), the human Caleb tests the AI robot Ava, but ultimately questions his own humanity, examines his eye for digital traces and cuts his skin to cause him to bleed.
Over the past 25 years of AI cinema, the boundaries between human and AI, digital and material, have become porous, highlighting the fluid and hybrid nature of AI creations. And in the films In The Machine (2013), Transcendence (2014) and Chappie (2015), the boundary between humans and AI is eroded almost to the point of non-existence. These films present scenarios of transhumanism – where humans can evolve beyond their current physical and mental limitations by harnessing the power of artificial intelligence to upload the human mind.
Although these stories are imaginary and their characters fictional, they vividly portray our fascinations and fears. We fear artificial intelligence and that fear never disappears in the film, even though it has been questioned more in recent decades and seen more positive portrayals, such as the little waste collector in WALL-E. But most of the time we fear that they will become too powerful and try to become our masters. Or we fear that they are hiding among us and that we may not recognize them.
But sometimes we also feel sympathy for them: AI characters in films can be pathetic figures who want to be accepted by humans, but never will be. We are also jealous of them – of their intellectual capabilities, their physical robustness and the fact that they do not experience human death.
Surrounding this fear and envy is a fascination with AI that is present throughout film history: we see ourselves in AI creations and project our emotions onto them. Sometimes enemies of humans, sometimes creepy mirrors, and sometimes even human-AI hybrids: the past seventy years of films about AI demonstrate the inextricably intertwined nature of human-AI relationships.
This article is republished from The Conversation under a Creative Commons license. Read the original article.
Paula Murphy does not work for, consult with, own shares in, or receive funding from any company or organization that would benefit from this article, and has disclosed no relevant affiliations beyond their academic appointment.