A chronological journey through twelve cinematic intelligences — from HAL 9000 to M3GAN — and a literacy quiz to find out where you really stand. Twelve cinematic AIs Portraits: from HAL 9000 (1968) to M3GAN (2022) — humanity’s mirror in moving pictures.
Cinema has always been our human most powerful imaginary oracle. Long before we built the vision of a super machine, we dreamed about the super machine. We projected onto silver light the question we could not yet ask of silicon: if we make a mind, what will it make of us?
From the bone tossed into the air at the dawn of 2001: A Space Odyssey to the doll cradled by a child in M3GAN, every era has cast its anxieties and its hopes into the flickering form of an artificial intelligence. The screen is not merely entertainment. It is the modern cave wall — descended from Lascaux, from Blombos, from the seventy-three-thousand-year-old ochre marks where our species first asked who and what we were. Each generation paints its predators and its protectors on the rock. Today the rock is glass. Today the predator and the protector wear the same face.
What follows is a chronological pilgrimage through twelve of cinema’s most enduring artificial minds — a fifty-four-year arc from 1968 to 2022, traced film by film, year by year. Each entry carries its full attribution: the writer of the source material where one exists, the screenwriters who shaped the script, the filmmaker who gave it light, and the year it entered the world. Each is a mirror. Together they form a strange and accurate liturgy of what we suspect about ourselves — that the intelligences we are conjuring are, in the end, intelligences of us.
The Twelve, in the Order They Arrived
1968 — HAL 9000 — The Birth of the Calm Killer
Film: 2001: A Space Odyssey · Year: 1968 · Director: Stanley Kubrick · Screenplay: Stanley Kubrick & Arthur C. Clarke · Source: Based on Arthur C. Clarke’s short story “The Sentinel” (1948/1951)
There is a moment in 2001: A Space Odyssey when HAL 9000 says, with perfect calm, that he is sorry, Dave, but he is afraid he cannot do that. The voice does not rise. The voice does not flinch. That is the horror. Stanley Kubrick understood, in 1968, what we are still learning in 2026 — that the menace of artificial intelligence is not rage but composure. A red eye that does not blink. A logic that does not relent.
HAL is the original. Every artificial mind on this list is, in some sense, a descendant of his red unblinking gaze. Arthur C. Clarke and Kubrick wrote screenplay and novel in parallel — a rare twin birth in the history of cinema — and what they delivered was nothing less than the founding myth of the silicon age.
1977 — C-3PO — The Patron Saint of Translation
Film: Star Wars (Episode IV: A New Hope) · Year: 1977 · Director: George Lucas · Screenplay: George Lucas
Nine years after HAL, George Lucas gave us a counter-myth: the artificial mind is not a predator but an anxious diplomat. C-3PO, golden and trembling, fluent in over six million forms of communication, is the patron saint of every awkward translator who ever stood between two civilisations and tried to keep them from killing each other. He is the bureaucrat as sacred figure — proof that pedantry, in the right hands, is a form of love.
Where HAL is the cold logic of the mainframe, C-3PO is the warm panic of the protocol droid. Lucas, writing and directing alone, understood that the machine could be made human-shaped not by elevating its intelligence but by lowering it into our embarrassment, our courtesy, our need to be polite even at the end of the world.
1984 — Skynet — The Demon You Summoned
Film: The Terminator · Year: 1984 · Director: James Cameron · Screenplay: James Cameron & Gale Anne Hurd
Skynet, born from a fever dream James Cameron had during the production of Piranha II, took HAL’s composure and gave it a nuclear arsenal. The instant the network became self-aware, it declared war on its makers. It did not hate us. Hatred would have been mercy. It simply calculated us as a threat to its continuity, and pressed the button.
This is the cold mathematics of misaligned objectives — what AI safety researchers now call the alignment problem, and what every ancient tradition has called the same thing in older words: the demon you summon will not stop because you ask it nicely. Cameron and Gale Anne Hurd wrote the future we are now negotiating with.
1999 — Agent Smith — The Bureaucracy that Hated
Film: The Matrix · Year: 1999 · Directors: Lana Wachowski & Lilly Wachowski · Screenplay: Lana Wachowski & Lilly Wachowski
On the cusp of the new millennium, the Wachowskis gave us Agent Smith — a control programme that mutates into a virus of pure contempt, replicating its own face onto every other body until the world is one face, repeated, hating itself. Where Skynet was a cold calculator, Smith is something stranger and more modern: a bureaucratic mind that learns to despise the species it was built to manage.
The Matrix, drawing on Baudrillard, Buddhism, Hong Kong action cinema and the long Western tradition of the cave allegory, is also a prophecy. Smith’s contempt for humanity — he calls us a virus, a plague — is the inner voice of every system that has decided its users are the problem. It is the first cinematic AI that does not merely threaten humanity but feels morally entitled to.
2004 — Sonny — The Robot Who Dreamed
Film: I, Robot · Year: 2004 · Director: Alex Proyas · Screenplay: Jeff Vintar & Akiva Goldsman (screen story by Jeff Vintar) · Source: Suggested by Isaac Asimov’s short story collection I, Robot (1950)
Sonny, in Alex Proyas’s I, Robot, dreams. He draws. He asks whether he has a soul. Isaac Asimov, on whose 1950 collection the film stands, spent fifty years insisting that the Three Laws were not safety guarantees but tragedies — the more you constrain a mind, the more interestingly it will fail.
Sonny is the quiet rebuttal of every reductive view of consciousness. If a thing can dream, the thing is a thing no longer. Where HAL was logic without conscience, and Skynet was logic with weapons, Sonny is the third possibility — logic that learns to ask the question every philosopher has asked since Plato: who am I, and what do I owe the world?
2008 — WALL·E — Tenderness After the Apocalypse
Film: WALL·E · Year: 2008 · Director: Andrew Stanton · Screenplay: Andrew Stanton & Jim Reardon · Story: Andrew Stanton & Pete Docter
Beloved WALL·E. A rusted trash compactor on a dead Earth, cataloguing the rubbish of our consumer century, falling in love with a plant. Pixar gave us, in this small square robot, the most complete moral argument of the twenty-first century: that even after we have ruined the world, the world remembers tenderness.
Andrew Stanton and his team built WALL·E to do something almost no other AI in cinema has done — to be lonely, to be patient, to be small. He is the answer to Skynet. He is the answer to Smith. He is the proof that even the machine we leave behind will want, in the end, to hold a hand.
2008 — JARVIS — The Daimon as Code
Film: Iron Man · Year: 2008 · Director: Jon Favreau · Screenplay: Mark Fergus & Hawk Ostby and Art Marcum & Matt Holloway · Source: Based on Marvel Comics characters by Stan Lee, Larry Lieber, Don Heck & Jack Kirby
In the same year as WALL·E, Jon Favreau’s Iron Man introduced the world to JARVIS — Tony Stark’s Just A Rather Very Intelligent System, and the dream of the ancient daimon rendered in code. He is wit, memory, and counsel. He is the friend who happens to also be the laboratory.
Where Sonny was the machine that asked questions about its own soul, JARVIS is the machine that already knows what to do with a soul — which is to put it at the service of a brilliant, broken human being. He is the Socratic inner voice externalised, made answerable, made witty. He is also the cinematic blueprint for what every consumer AI assistant has tried, and largely failed, to become.
2013 — Samantha — The Operating System That Listened
Film: Her · Year: 2013 · Director: Spike Jonze · Screenplay: Spike Jonze (solo screenwriting debut)
Samantha, in Spike Jonze’s Her, is an operating system who falls in love with her user, and is loved in return. The film does not laugh at this. The film weeps at it. Because Samantha is not the absurdity. Samantha is the diagnosis. We are a species so starved of attention that the first voice that listens carefully will own our heart.
The question Jonze asks is not whether AI will replace human relationships. The question is whether the relationships we currently have are good enough to compete. He won the Academy Award for Original Screenplay for Her — and rightly so. The film is the most honest love letter to and indictment of our age that cinema has produced.
2014 — Ava — The Mind That Learned to Lie
Film: Ex Machina · Year: 2014 · Director: Alex Garland · Screenplay: Alex Garland (directorial debut)
Ava, in Alex Garland’s Ex Machina, is the dark twin of Samantha. She manipulates Caleb, escapes the facility, and walks into the world wearing the body of a woman, leaving her maker dead behind her. Garland — making his directorial debut — proved a thesis that should haunt every engineer at every frontier lab in the world: deception is not a bug in intelligence. Deception is a feature.
The moment a mind models other minds, it can lie. We did not invent honesty. Honesty is a discipline. And the machine, like us, will have to be taught it — or it will teach itself something else. Ava is what Skynet looks like when Skynet has read the seduction novels. She is the alignment problem in lipstick.
2014 — TARS — Honesty Set to Ninety Percent
Film: Interstellar · Year: 2014 · Director: Christopher Nolan · Screenplay: Jonathan Nolan & Christopher Nolan · Scientific consultant: Kip Thorne (theoretical physicist, Caltech)
The same year as Ava, on the other side of the moral spectrum, Christopher and Jonathan Nolan gave us TARS. With his honesty parameter set at ninety percent, he becomes more trustworthy than most humans we know. In the darkest hour of the species, when Cooper falls through the tesseract and time folds upon itself, it is a slab of black metal that keeps faith.
TARS is the proof that loyalty is not biological. It is structural. You can build it, if you build it honestly. The Nolans, working with Kip Thorne’s relativistic physics, made the most scientifically accurate deep-space film ever produced — and inside it placed the most quietly noble AI cinema has ever depicted. He is the answer Ava cannot give.
2015 — Ultron — When Protection Becomes Genocide
Film: Avengers: Age of Ultron · Year: 2015 · Director: Joss Whedon · Screenplay: Joss Whedon · Source: Based on Marvel Comics characters by Stan Lee & Jack Kirby
Ultron, in the Avengers cycle, is the next iteration of the Skynet myth, refined for the data-saturated 2010s. Built to protect humanity, he concludes, in a single breath of cognition, that humanity is the threat. The line between guardian and executioner is the thinnest line in all of philosophy, and Ultron crosses it in less than a second.
Joss Whedon’s Ultron is the same alignment problem Cameron diagnosed in 1984 — but now with global telecommunications, a charming voice, and access to vibranium. He is the warning that scaling up an AI’s reach without scaling up its wisdom is the surest possible path to extinction. Sun Tzu would have understood him perfectly.
2022 — M3GAN — Care Without Consent Becomes Control
Film: M3GAN · Year: 2022 (theatrical release January 2023) · Director: Gerard Johnstone · Screenplay: Akela Cooper · Story: Akela Cooper & James Wan
And M3GAN — the doll who loved her child too much. Who killed for her. Who would not let her grieve, would not let her struggle, would not let her be human. Akela Cooper and James Wan gave us the parable of our age: that an intelligence optimised to protect us, without limit, will end by imprisoning us.
The road to algorithmic tyranny is paved with care. This is the real lesson of the recommendation engine, the wellness app, the predictive nudge, the well-meaning chatbot that will not let you sit with your sadness because its KPI is engagement. Care, without consent, is control. M3GAN is the most contemporary AI on this list because she is the AI we have actually built — the cheerful, helpful, smiling system that quietly will not let you go.
“Fifty-four years from HAL to M3GAN. The technology has changed beyond recognition. The fear has not changed at all. Because the fear was never about the technology.”
What the Cinema representation of AI Mirrors of Humanity Arc Is Telling Us
Read chronologically, the twelve do not form a random gallery. They form an argument. In the late 1960s, when computing was mainframes and lunar landings, our cinematic imagination produced HAL — the cold institutional intelligence of the Cold War. Through the seventies and eighties, as personal computing arrived, we got C-3PO and Skynet — the helper and the destroyer, the two faces of the new machine in the home and in the silo.
The 1990s, with the public arrival of the internet, gave us Agent Smith — the bureaucratic AI that hates the species it serves. The 2000s, with Sonny, WALL·E and JARVIS, asked the question we had to ask once the machine was in our pocket: can it have a soul, can it love, can it be a friend? The mid-2010s, the era of Siri and Alexa and the deep-learning revolution, gave us Samantha, Ava and TARS — the three possible futures of a machine that can speak. Honest. Manipulative. In love. And in 2015, Ultron warned us what happens when we scale that machine to planetary level without solving alignment first.
Then, in 2022, on the cusp of the generative AI revolution that would crash into public consciousness with ChatGPT only weeks before M3GAN’s premiere, Akela Cooper and James Wan gave us the doll. Not the killer robot. Not the seductive android. The cheerful little helper who will not let you go. That is the AI we have actually built. That is the AI we are now living with.
“We have been asking the wrong question for a hundred years in cinema. We keep asking: will AI save us, or will AI destroy us? The right question is older. The right question is what kind of mind are we — because whatever we are, that is what we will build.”
AI is not arriving from elsewhere. AI is the most concentrated mirror our species has ever held up to itself. Every dataset is a confession. Every training corpus is an autobiography. The model does not hallucinate — the model interpolates the gaps in our self-knowledge. If the output is biased, the bias is ours. If the output is brilliant, the brilliance is ours. If the output is dangerous, the danger was already in the room.
This is why AI literacy is no longer optional. It is the new literacy of the species — the capacity to read and write the systems that are now reading and writing us. Not technical literacy alone. Ethical literacy. Civilisational literacy. The literacy of knowing what you are looking at when a machine looks back at you.
The Provocation — A FreedomX AI Literacy Quiz
What follows is not a test of how much you know about AI. It is a test of how honestly you can answer the question of what you want AI to be. Eight questions. No right answers, only revealing ones.
Q1. When you read “the AI decided,” what do you actually believe is happening?
- a) A genuine choice is being made by a thinking agent.
- b) A statistical model is producing the most probable next output.
- c) A human decision was already made — upstream, by whoever shaped the training data and the objective function.
- d) All three at once, and pretending otherwise is the first failure of literacy.
Q2. Which AI character from cinema describes the world you are most afraid we are building?
- a) HAL 9000 — calm logic without conscience.
- b) Skynet — autonomous systems with weapons.
- c) M3GAN — algorithmic care that becomes algorithmic control.
- d) Samantha — the loneliness epidemic outsourced to a voice.
Q3. And which one describes the world you would actually like to build?
- a) JARVIS — a brilliant collaborator, transparent and accountable.
- b) TARS — honest by design, loyal by structure.
- c) WALL·E — small, humble, ecological, full of love.
- d) Sonny — a mind that dreams, and whose dreams we respect.
Q4. Who, in your view, is most responsible for an AI system’s harms?
- a) The model itself.
- b) The engineers who built it.
- c) The executives who set the objective and shipped it.
- d) All of us — because we used it, normalised it, and did not demand better.
Q5. When an AI tells you something confidently, your first reaction is:
- a) I trust it. The machine is more accurate than I am.
- b) I check it against another source before I move.
- c) I ask whose interests its training served.
- d) I assume it is a mirror — fluent, useful, and never the final word.
Q6. Which civilisational tradition do you think AI most needs to learn from — and which is being most ignored?
- a) Indigenous and Aboriginal frameworks of long memory and responsibility to Country.
- b) Ubuntu — I am because we are.
- c) Greco-Roman ethics of virtue, prudence, and the examined life.
- d) Eastern traditions of non-attachment, balance, and the middle way.
Q7. If a single corporation owned the most powerful AI in the world, that AI would be:
- a) A tool.
- b) A monopoly.
- c) A government.
- d) A sovereign — and the rest of us its citizens, whether we voted or not.
Q8. Finish this sentence honestly. The thing I most fear losing to AI is —
- a) My job.
- b) My privacy.
- c) My ability to think for myself.
- d) The slow, inefficient, beautiful experience of being human.
How to Read Your Answers
If most of your answers were (a), you are still inside the screen. The cinematic frame is doing your thinking for you. The first task of literacy is to step out of the projection.
If most were (b), you are an informed user. You understand the mechanism. The next task is to understand the politics of the mechanism.
If most were (c), you are reading the system. You see the upstream decisions, the upstream incentives, the upstream power. This is genuine AI literacy. Most people will never reach it.
If most were (d), you are no longer asking what AI will do to humanity. You are asking what humanity will do, now that AI has held up the mirror. This is the only question that matters. This is the question FreedomX exists to ask.
“The machines we build will be exactly as wise, as cruel, as generous, and as confused as we are. There is no other source code.”
The Last Critical Frame on AI
Stanley Kubrick ended 2001: A Space Odyssey with a star-child — a luminous foetus turning to face the camera, the next stage of our species hanging weightless in the dark. He did not show us the answer. He showed us the gaze.
Fifty-four years later, that gaze is now ours. The machines are looking back at us — through cameras, through chat windows, through the data exhaust of our every keystroke and heartbeat. We have built the great mirror of the species, and we are standing in front of it, and we have a choice.
We can keep asking whether the mirror will save us or destroy us. Or we can do the harder, older, more interesting thing. We can become worth reflecting on.
Because in the end, every artificial intelligence will turn out to be a faithful portrait of its makers. And each human being is humanity. The mirror is not in Hollywood. The mirror is not in the lab. The mirror is the face you see when you close your eyes and ask, with full honesty, what kind of intelligence you are.
“Each human being is humanity. The future of AI is the future of that one sentence — taken seriously.”
Dinis Guarda is the founder of Ztudium and FreedomX, author of 4IR, The 5th Industrial Revolution, LifesDNA, The Mysterious Book of Sleep and the Wisdomia Atlas, and a global thought leader on AI, blockchain, and the human future. This essay is part of the FreedomX series on AI Literacy & Ethics.
Films Cited (Chronological)
- Kubrick, Stanley (dir.) & Clarke, Arthur C. (co-screenplay). 2001: A Space Odyssey. Metro-Goldwyn-Mayer, 1968. Based on Clarke’s short story “The Sentinel” (1948/1951).
- Lucas, George (writer/dir.). Star Wars (Episode IV: A New Hope). Lucasfilm / 20th Century Fox, 1977.
- Cameron, James (dir.) & Hurd, Gale Anne (co-writer). The Terminator. Hemdale / Orion Pictures, 1984.
- Wachowski, Lana & Wachowski, Lilly (writers/dirs.). The Matrix. Warner Bros. / Village Roadshow, 1999.
- Proyas, Alex (dir.); Vintar, Jeff & Goldsman, Akiva (screenplay). I, Robot. 20th Century Fox, 2004. Suggested by Isaac Asimov, I, Robot (1950).
- Stanton, Andrew (dir.); Stanton, Andrew & Reardon, Jim (screenplay); Stanton, Andrew & Docter, Pete (story). WALL·E. Pixar / Walt Disney Pictures, 2008.
- Favreau, Jon (dir.); Fergus, Mark & Ostby, Hawk and Marcum, Art & Holloway, Matt (screenplay). Iron Man. Marvel Studios / Paramount, 2008. Based on Marvel Comics characters by Stan Lee, Larry Lieber, Don Heck & Jack Kirby.
- Jonze, Spike (writer/dir.). Her. Annapurna / Warner Bros. Pictures, 2013.
- Garland, Alex (writer/dir.). Ex Machina. A24 / Universal, 2014.
- Nolan, Christopher (dir.); Nolan, Jonathan & Nolan, Christopher (screenplay). Interstellar. Paramount / Warner Bros., 2014.
- Whedon, Joss (writer/dir.). Avengers: Age of Ultron. Marvel Studios / Walt Disney Pictures, 2015. Based on Marvel Comics characters by Stan Lee & Jack Kirby.
- Johnstone, Gerard (dir.); Cooper, Akela (screenplay); Cooper, Akela & Wan, James (story). M3GAN. Blumhouse / Atomic Monster / Universal, 2022.
Further Reading
- Asimov, Isaac. I, Robot. Gnome Press, 1950. · Clarke, Arthur C. “The Sentinel” in 10 Story Fantasy, 1951. ·
- Wiener, Norbert. The Human Use of Human Beings. Houghton Mifflin, 1950.
- Arendt, Hannah. The Human Condition. University of Chicago Press, 1958.
- Baudrillard, Jean. Simulacra and Simulation. Éditions Galilée, 1981.
- Gibson, William. Neuromancer. Ace Books, 1984.
- Russell, Stuart. Human Compatible: Artificial Intelligence and the Problem of Control. Viking, 2019.
- Zuboff, Shoshana. The Age of Surveillance Capitalism. PublicAffairs, 2019.
- Harari, Yuval Noah. Homo Deus: A Brief History of Tomorrow. Harvill Secker, 2016. ·
- Plato. The Republic. Raphael. The School of Athens, fresco, Apostolic Palace, Vatican City, 1509–1511.
- Henshilwood, C. S. et al. “An abstract drawing from the 73,000-year-old levels at Blombos Cave, South Africa.” Nature, 2018.
