Probably one of the most chilling scenes in film history is the deactivation of HAL 9000 in 2001: A Space Odyssey. It's the culmination of the film's central episode, the journey of the spaceship Discovery to Jupiter. Its crew consists of five astronauts, three of whom are in hibernation while the other two navigate in a peacefully tedious atmosphere, because the control functions are performed by a computer called HAL 9000. At the beginning of the episode, a clever narrative idea (not in the original script)1 introduces us to the mission's characters through a pre-recorded television program that the crew watches. In it, the television announcer interviews HAL (because HAL speaks with an eerily calm voice, if you'll pardon the oxymoron) and explains to the viewers that it possesses "Machine Intelligence," an expression that the Spanish translator rendered as "inteligencia mecánica" (mechanical intelligence) in the dubbed version released in 1968; significantly, the film's current subtitles translate it as "artificial intelligence." This expression already existed when the film was made, because in 1956 it was introduced by John McCarthy to label computer efforts to simulate or reproduce certain aspects of human intelligence; but it was not, evidently, as widespread as it is now. Perhaps it could have been translated as "machinic intelligence," a somewhat unusual term, but one that more accurately describes the reality of a supposedly intelligent machine; another possibility would be to translate it as "automatic intelligence," just as is done today with "machine learning," which has been translated as "aprendizaje automática" (automatic learning). Different nuances, since, as the Italians well know, "traduttore traditore" (translator, traitor). In any case, Stanley Kubrick wrote the screenplay with Arthur C. Clarke, a science fiction writer and popularizer of science who undoubtedly knew what he was talking about.
What's important for the film's story is that HAL 9000 is endowed with what we now call Artificial Intelligence. And it so happens that, in its duties controlling the ship, it detects a fault in a transmission antenna; checks carried out by the human crew show that there is no such fault, but HAL continues to insist that there is a malfunction. Faced with this unprecedented situation (nothing like this has ever happened with HAL 9000), the crew considers shutting down the computer. HAL discovers these intentions and, to prevent it, is determined to eliminate all the human crew members. It almost succeeds, but one survives, Dave Bowman, who ultimately shuts it down. These humble words of mine fail to even begin to capture the mastery with which Kubrick tells the story. As we listen to Dave Bowman's labored breathing inside his helmet and observe his movements within shots constructed with obsessive symmetry (a Kubrick trademark), HAL intervenes, asking Bowman to stop and talk. The astronaut follows his instructions with painstaking precision until, at a certain point, HAL says, “I'm scared, Dave”. I don't know if it was Kubrick or Clarke who introduced that line (which wasn't in the initial shooting script), but in it they encapsulated the key to Artificial Intelligence. Because, with those words, HAL demonstrates an awareness of its individual existence, its finitude, and the unfathomable and terrifying abyss that this awareness of finitude creates. And it shows, therefore, that true intelligence, human intelligence, the only kind we know through experience, doesn't consist solely of processing data or manipulating logical operations.
However, HAL's workings remain enigmatic. Perhaps, given that in fiction it is indeed a genuine intelligence, we could speak of action or behavior on the part of this computer. Although I have seen the film several times (and it has always fascinated me with the same intensity), I have never fully understood the reason for HAL's rebellion. This ambiguity is not surprising. In making the film, Kubrick opted for a deliberately open and metaphorical tone, leaving the search for and discovery of possible meanings in the hands of the viewer; and, at times, Kubrick didn't even leave hidden meanings, but rather directly unfolded the images for the free enjoyment and creative intuition of the viewer. In this way, Kubrick openly distanced himself from the approach of his collaborator Clarke, who favored scientific explanations to clarify the narrative.2 But, regardless of stylistic and narrative choices, placing HAL's fear at the key moment of the scene demonstrates Kubrick's depth and perspicacity. His interest in artificial intelligence is shown by his project in the 1990s to film a movie based on a story by Brian Aldiss, The Supertoys. The project lasted all summer and centered on a robot in the form of a child; it was finally directed by Steven Spielberg after the death of his friend Kubrick.3
It is precisely this motivational ambiguity that acutely reflects the problems of designing human intelligence through the intricacies of computer technology. Faced with a complex situation encompassing a secret mission still unknown to the human crew, potential technical failures, and interaction with humans who are considering disconnecting, HAL must make decisions, weigh options, and, above all, make moral choices; this last point is fundamental: because, faced with the prospect of failure, the computer decides to sacrifice the humans. It is reasonable to think that this computer, despite its organizational capacity, lacks a moral conscience and, consequently, considers human lives just another factor influencing the mission and expendable for the sake of its success; HAL is aware that this success inevitably requires the capabilities that only it can deploy. It is also worth considering the hypothesis (which does not exclude the above) of a self-aware HAL that prioritizes its survival against disconnection (the death of a computer) above all else, even the lives of the human crew. Violence, death, and the struggle for survival would be considered part of human intelligence for Kubrick if we consider the film's first episode: Dawn of Man. Primates, expelled from a watering hole by other primates after discovering a black monolith, learn to use weapons to kill animals and eliminate their enemies. A similar (and larger) monolith awaits the mission on Jupiter, a fact known only to HAL aboard the ship.
All these motives, perceptions, and feelings converge in the moment when HAL acknowledges its fear. Because what we call intelligence is a complex system irreducible to fixed rules, structured not only by formal laws, which we call logic, forming a framework of rationality, but which only explain a portion of the processes and intricacies of intelligence. This complexity, depicted in the film, is truly unreproducible through computer programs, and therefore we only have what engineers call weak AI, incapable of self-awareness or of performing multiple tasks and learning processes. We don't know how to move forward in this direction, probably because the gap between programming embedded in silicon and genuine intelligence implies the existence of qualities beyond the reach of computing. We could program a machine so that, upon receiving a certain input, it would utter the phrase: I am afraid. But we cannot program a machine to feel fear.
1 Cfr. https://www.archiviokubrick.it/opere/film/2001/script/2001-originalscript.pdf, accessed on 4-08-2026.
2 Cfr. Ciment, M., Kubrick. Owl Book, New York, 1984, pp. 96-97.
3 Cfr. Kolker, R.P. / Abrams, N., Kubrick. An Odyssey. Faber, London, 2024, p. 257.
