If the 1950s had more than their fair share of robots, the 1960s were an age of cinematic computers. Unlike robots, computers found themselves appearing outside of science fiction-themed films. To be fair to the 1950s, that trend was apparent as early as 1957 when a business computer was central to the plot of the Spencer Tracy/Katharine Hepburn vehicle The Desk Set. Computers and romantic comedy are not an obvious match, but The Desk Set touches on the broader theme of computers/robots displacing people from their jobs. Specifically, Spencer Tracy plays a character installing a computer that threatens to displace Katharine Hepburn’s role as a librarian. It all ends happily, as you might expect.
An even more unlikely computer-themed romantic comedy is 1961’s The Honeymoon Machine, which somehow stars Steve McQueen. I haven’t seen the film, so I only have a plot synopsis to go off, but it involves sailors using a US Navy computer to win at roulette at a casino. It seems unlikely to me that this is the first notable fictional treatment of using computers to gain an advantage over a casino, but I’m not aware of an earlier film treatment of the idea.
Moving closer to science fiction is the 1967 spy thriller Billion Dollar Brain. The third and final adaptation of Len Deighton’s Harry Palmer 1 stories, starring Michael Caine as Palmer. The plot involves an anti-Communist Texan billionaire who has used his wealth to build a computer that helps him manage a large network of Latvian agents.
In all three cases, the computer’s capability is likely beyond what was possible at the time, and so, arguably, all three are science fiction. However, each of the computers is meant to be contemporary technology and not artificial intelligence as such (despite them being referred to as brains). A partial distinction from robot stories is how the computers act as a kind of oracle, anticipating events and guiding action (not always correctly).
Stepping even further back towards science fiction is Jean-Luc Godard’s dystopian 1965 detective noir film Alphaville. Dour trench-coat-wearing detective Lemmy Caution2 investigates the city of Alphaville, in what is implied to be a sort of future version of 1960s France, i.e. the setting is not presented as futuristic 3, but textually the setting isn’t the present, and the dialogue is peppered with oddly placed science-like references to other galaxies and light-years.
Alphaville is a city in which individualism has been undermined, along with emotion and artistic expression. Ostensibly to create a more rational society, the city is under the control of a central computer called Alpha 60, which maintains order through brainwashing and murder.
The computer, as a symbol for a brutal, rationalist re-ordering of society, reaches a peak in the 1970 film Colossus: The Forbin Project. The US government builds a secret computer in an underground facility with its own power plant. The computer (Colossus4) is tasked with controlling the US nuclear deterrent. Once activated, Colossus deduces the existence of an equivalent Soviet computer. Inevitably, the two computers collaborate to take control of the world, surprisingly successfully.
Colossus is, of course, another variant on the Frankenstein narrative, with the inventor Charles Forbin in the Victor role and the computer as the creature. The name itself is a reference to larger-than-life statues, which is a further echo back to Talos of Greek myth. In the original novel by D.F.Jones, there are overt references to Mary Shelley’s novel.
“Fisher glanced round, gave his tie another tug, and lowered his voice as he moved closer. ‘To be honest, I rather wish these people weren’t so cocksure and happy. Perhaps it’s because we are practically out of a job, or it’s anti-climax — but I think I have allowed for all that. Yet I’m left with a nasty taste, and I can’t really say what it is.’ ‘Is it Colossus?’
Jones, D. F.. Colossus (pp. 39-40). (Function). Kindle Edition.
‘Yes, I guess it is,’ Fisher gave a short false laugh. ‘Frankenstein should be banned reading for scientists.’
‘I’d be more inclined to make it compulsory reading for non-scientists’ Forbin peered into his glass.”
Later, the computer is referred to as “Frankenstein’s monster himself”. The analogy is imperfect, as Colossus happily follows his purpose, just in a way that his creators did not anticipate. Instead, we have the idea of Frankenstein as a morality tale about scientific hubris.
As an example of nuclear war thriller, Colossus, owes a little to Stanley Kubrick’s dark comedy Doctor Strangelove. Computers aren’t central to the plot of Doctor Strangelove, but automated systems intended to circumvent human error do play a part in ensuring the eventual doom of humanity. I feel Strangelove better understands that the nuclear detterent is to root of the insanity than Colossus does. Then again, I first saw the film in the 1980s when Ronald Reagan was US President, and so world computer dictatorship felt like a better option than nuclear annihaltion. These days, I suppose the choice would be between Donald Trump and Grok, so the choices have only got worse.
I’ve saved Hal, the computer featured in Stanley Kubrick’s 2001: A Space Odyssey, until last, even though Hal predates Colossus by a couple of years. Partly because I discussed whether Hal counts as a robot a few years ago, but now I need to put him back in the category of computers for the purpose of this chapter.
As with Alpha 60 and Colossus, Hal is both a computer and a character within the story. Hal is also an example of a sentient computer being placed in a position of power over people, with murderous consequences.
Hal is the most sympathetic of the trio of computer tyrants. Perhaps my perspective is shaped by the novel and sequel. The film provides little reason why Hal acts against the crew of the Discovery One but the implication is that it is a mix of malfunction and a desire for self-preservation. Additionally, Hal is aware of the secret aspect of the mission to Jupiter (to find an ancient monolith), which the crew is not.
2001 mixes aspects of “hard” science fiction with more fantastical elements. However, Hal is clearly intended to be more on the “likely future progress” than the trippy-alien-space-gods end of the spectrum. By the actual year 2001, the apparently sapient Hal looked absurd as prediction of how AI would progress from the late 1960s. Computers had advanced tremendously, but the most recent major issue of with computers had been the potential for disaster in the year 2000 if older software got confused by the date.
While I don’t believe we are much closer to computers with genuine human-like intelligence, it is also true that what we actual see Hal do (as opposed to any mental states we might imagine it has) is not far from current technology. Hal can manage natural language and autonmously control other computerised systems on the ship. There would be a host of engineering problems getting such a computer to work on a spaceship but then again there would be a host of engineering problems making a spaceship like Discovery One. Arguably, we are closer to making Hal than we are to making a spaceship to take people to Jupiter.
Of course, we also would not be surprised now that if we put ChatGPT in charge of a spaceship that it might “decide” to kill the crew as the best way of completing a mission. Yet also, we would not be surprised when it turned out some company had decided to put ChatGPT (or similar) in charge of a situation where it might kill people in the belief it was the best way of completing a mission.
- At least, Harry Palmer is the agent’s name in the films – I think in the books, the agent doesn’t have a name ↩︎
- Played by Eddie Constantine – a man whose actual name also sounds like the kind of character he plays. ↩︎
- It was filmed in Paris but features modernist buildings. ↩︎
- Elon Musk would later use this as the name of the supercomputer running the Grok large language model. ↩︎
Camestros Felapton 