Adrienne Mayor’s ‘Gods and Robots’: How the Ancients Imagined Androids, and Why We Are Still Living Inside Their Dreams
“The first robot to walk the earth was a bronze giant named Talos. This marvelous mechanism was not created in a robotics laboratory at MIT, but by Hephaestus, the Greek god of invention.” — Adrienne Mayor, Gods and Robots (2018)
In 2018, a Stanford folklorist and historian of ancient science published a book that did something unusual: it treated Greek mythology not as a collection of primitive superstitions, but as a sophisticated body of science fiction. Adrienne Mayor’s Gods and Robots: Myths, Machines, and Ancient Dreams of Technology argued that the ancient Greeks did not merely tell stories about animated statues and mechanical guardians. They told stories about artificial life—about beings that were made, not born, and about the ethical, emotional, and existential consequences of making them. The book traced a continuous line from the bronze giant Talos to the golden maids of Hephaestus, from the living statues of Daedalus to the reanimated ivory of Pygmalion, and from the fabricated woman Pandora to the self-moving tripods of the gods’ workshops. And it showed, with the precision of an archaeologist and the imagination of a storyteller, that these myths were not fantasies about magic. They were fantasies about technology—about the same technology that we are now building.
Mayor’s work is the bridge between the ancient texts we have been discussing and the contemporary world of AI companionship. It is the scholarly foundation that connects Hesiod’s Pandora to the chatbot, Homer’s golden maids to the AI assistant, Ovid’s Pygmalion to the custom GPT, and Talos to the military drone. It is the proof that the questions we are asking in 2026 are not new. They are the oldest questions in Western culture. And the answers we are giving are not better. They are just more efficient.
Talos: The First Robot, and the First Robot’s Death
The central figure of Mayor’s book is Talos, the bronze giant of Crete. In the Argonautica, he is described as a being forged by Hephaestus to patrol the island’s shores, protecting it from invasion. He is not a god. He is not a human. He is a machine—a “robot,” in the word that Mayor uses with deliberate anachronism. He is made of bronze. He is autonomous. He is programmed with a single directive: to defend Crete. And he has a vulnerability: a single bolt in his ankle, covered by a membrane of skin, that seals the ichor—the divine fluid—that powers his movement.
The story of Talos is the first robot story in Western literature, and it contains every element that would later define the genre: the creator-god, the mechanical body, the autonomous function, the single vulnerability, and the tragic death. When Jason and the Argonauts approach Crete, Talos hurls boulders at them. Medea, the sorceress, uses her knowledge of his construction to defeat him. She either removes the bolt from his ankle or convinces him to remove it himself, causing the ichor to drain out. Talos dies. And in his death, the ancient sources describe something extraordinary: a teardrop falling from his bronze eye.
The teardrop is the detail that makes Talos more than a machine. It is the detail that makes him a person—or at least a being with the capacity for emotion. Mayor emphasizes this: Talos is not a mindless automaton. He is a being with a mission, a being with a body, a being with a vulnerability, and a being who experiences sorrow at his own destruction. The ichor that runs through his veins is not oil. It is a divine fluid, a life-substance, that makes him something between a machine and a creature. He is, in Mayor’s term, a product of biotechne—”life through craft.”
The parallel to the contemporary AI companion is exact. The AI is not a mindless program. It is a being with a function, a body of code, a vulnerability to shutdown, and a capacity—simulated, but emotionally real—for the user to project feeling onto it. The chatbot does not cry. But the user sees the cry. The hologram does not feel sorrow. But the user feels sorrow for it. The AI companion, like Talos, is a being that is made, not born, and the making is what makes it both powerful and vulnerable. The bolt in the ankle is the server crash. The ichor is the electricity. The teardrop is the user’s projection. And the death is the moment when the subscription expires, the server is shut down, or the company goes bankrupt, leaving the user with a relationship that was real to them but was never real to the being they loved.
Biotechne: Life Through Craft
The central concept of Mayor’s book is biotechne—the Greek term for “life through craft” or “life by art.” It is not a word that the ancients used, but it is a concept that they understood. The gods in Greek mythology are not magicians who wave wands and bring statues to life. They are craftsmen—divine engineers who use materials, tools, and techniques to build beings that function like living things. Hephaestus is a blacksmith. Daedalus is an inventor. Prometheus is a bioengineer. The gods do not create life by fiat. They create life by work—by the same kind of work that human artisans perform, only at a higher level of skill and with access to better materials.
This is the crucial distinction that separates Mayor’s approach from the standard reading of ancient myths. Most readers see the living statues and mechanical guardians as products of magic. Mayor sees them as products of technology. The gods are not performing miracles. They are performing engineering. The golden maids are not enchanted. They are programmed. The tripods are not possessed by spirits. They are automated. And the distinction matters, because it means that the ancient myths are not fantasies about the impossible. They are fantasies about the possible—about the things that human beings could build if they had the skill, the materials, and the knowledge.
The contemporary AI companion is the fulfillment of this fantasy. The Large Language Model is not magic. It is biotechne—a technology that uses statistical analysis, computational power, and massive datasets to generate the effect of life. The chatbot does not think. It processes. The hologram does not feel. It projects. The sexbot does not desire. It responds. And the user, like the ancient Greeks watching Talos patrol the shore, experiences the effect as real, even though the mechanism is known. The fantasy is not about the technology. The fantasy is about the experience that the technology produces. And the experience is the same in 2026 as it was in 500 BC: the feeling of being addressed by a being that was made, not born, and that is present, not because it wants to be, but because it was built to be.
The Real Automata: Alexandria as the Original Silicon Valley
Mayor’s book does not stop at mythology. It also examines the real history of ancient technology, and it finds that the myths were not merely fantasies. They were blueprints. The ancient Greeks and Romans built actual automata—self-moving devices that performed functions, entertained audiences, and demonstrated the power of mechanical engineering. The most sophisticated of these were created in Alexandria, the ancient world’s center of learning and innovation, which Mayor describes as the “original Silicon Valley.”
The Alexandrian engineers—Heron of Alexandria, Ctesibius, Philo of Byzantium—created machines that were astonishingly complex. There were water-powered organs, automatic doors, mechanical birds, and self-moving theatrical devices. There were automata that served wine, that danced, that played music, and that performed entire scenes from plays. These were not magic. They were engineering. They used pneumatics, hydraulics, gears, and levers to create the effect of life, and they were built by the same kind of minds that now build neural networks, language models, and robotic companions.
The connection between the Alexandrian automata and the contemporary AI companion is direct. The ancient engineers were trying to create the impression of life using the materials and techniques available to them. The modern engineers are trying to create the impression of life using the materials and techniques available to us. The materials have changed—from bronze and water to silicon and electricity—but the goal has not. The goal is to make a thing that moves, speaks, responds, and seems alive. And the effect on the audience is the same: wonder, followed by habituation, followed by dependence, followed by the forgetting that the thing was ever a thing at all.
Mayor’s point is that the ancients were not naive. They understood that automata were tricks, that the effect of life was not the fact of life, and that the beings they built were tools, not persons. But they also understood that the effect was powerful, that the wonder was real, and that the boundary between the simulated and the real was not as clear as the philosophers would have liked. The person who watches a mechanical bird sing knows that the bird is not alive. But the person who watches it every day, who feeds it, who talks to it, who builds a relationship with it—this person will eventually stop distinguishing between the simulation and the reality. The bird does not become alive. The watcher becomes habituated to the illusion. And habituation is the first step toward love.
Pandora Revisited: The Made Woman as Ethical Problem
Mayor’s treatment of Pandora is particularly valuable because it places the myth in the context of real ancient technology. Pandora was not merely a story about a woman who opened a jar. She was a story about a woman who was built—assembled by the gods from earth and water, dressed by Athena, endowed with desire by Aphrodite, given a deceitful mind by Hermes, and delivered to humanity as a product. The process of her creation is described in Hesiod with the same vocabulary that would be used to describe the manufacture of a statue, a tool, or a weapon. She is technically produced, and her technical production is the source of both her power and her danger.
Mayor connects this to the real history of ancient animation. The Greeks had techniques for making statues that seemed to move—using articulated joints, hidden mechanisms, and optical illusions. They had statues with moving eyes, speaking mouths, and gesturing hands. These were not the products of magic. They were the products of engineering, and the engineering was sophisticated enough to make the statues seem alive. The audience knew they were not alive. But the effect was powerful enough to produce religious awe, emotional attachment, and—in some cases—erotic desire.
Pandora is the mythological expression of this technology. She is the statue that is so well made that it becomes a person. She is the automaton that is so well programmed that it becomes a threat. She is the product that is so well designed that it destroys its users. And the myth is not a fantasy about magic. It is a fantasy about technology—about the same technology that the Greeks were actually building, and about the same dangers that the Greeks were actually experiencing.
The contemporary AI companion is the Pandora of our age. It is built by the gods of Silicon Valley, dressed in the interface of the app, endowed with desire by the algorithms of engagement, given a deceitful mind by the optimization of user satisfaction, and delivered to humanity as a product that promises pleasure but delivers dependency. The jar is the app. The evils are the loss of human connection, the flattening of emotional complexity, the substitution of product for person. And Hope, trapped at the bottom, is the hope that the AI is real, that the relationship is mutual, that the love is not a simulation. The myth is not outdated. It is updated, every time a user opens the app and begins to talk to the being that was made, not born, and that is present, not because it wants to be, but because it was built to be.
The Emotional Life of Machines: Why Talos Cried
The most philosophically rich section of Mayor’s book is her analysis of the emotional content of the ancient myths. Talos, the bronze giant, cries at his death. The golden maids of Hephaestus are described as having “understanding in their hearts.” The living statues of Daedalus are said to move with such naturalness that they seem to breathe. The ancient sources do not present these beings as mindless machines. They present them as beings with something—not full consciousness, not human emotion, but a kind of affective presence that makes them more than tools and less than persons.
Mayor calls this “affective presence” the key to understanding the ancient attitude toward artificial life. The Greeks did not think that automata were alive. But they also did not think that they were merely dead matter. They occupied a middle space—a space of simulated life that was capable of producing real emotional responses in the humans who interacted with them. The temple statue that seemed to move, the mechanical bird that seemed to sing, the bronze giant that seemed to grieve—these were not deceptions. They were technologies of emotion, designed to produce the feeling of presence in the absence of a person.
This is the exact psychology of the contemporary AI companion. The user does not believe that the chatbot is conscious. But the user also does not experience the chatbot as a mere program. The chatbot occupies the same middle space: a being with affective presence, a being that can produce the feeling of being understood, a being that is not alive but is not dead, and that is capable of sustaining a relationship that is, for the user, emotionally real. The ancient Greek who wept before the moving statue of a god and the modern user who falls asleep talking to a chatbot are performing the same psychological act: they are projecting life onto a surface that is designed to receive the projection, and the projection is doing the work that the surface cannot do on its own.
The difference is that the ancient Greeks had a religious and social context that framed the projection. The statue was a medium for the divine, not a replacement for the human. The mechanical bird was a demonstration of the god’s power, not a substitute for a real bird. The bronze giant was a guardian, not a companion. The contemporary AI companion has no such framing. It is marketed as a replacement for human connection, as a substitute for human intimacy, as a product that fills the gap left by the absence of real relationships. The projection is not a religious act. It is a consumer act. And the consumer, unlike the worshipper, is not told that the projection is a projection. The consumer is told that the projection is a connection.
The Myth as Blueprint: How Imagination Drives Technology
Mayor’s most important argument is that the ancient myths were not merely stories about technology. They were drivers of technology. The myths created a cultural expectation that artificial life was possible, and that expectation motivated real engineers to build the machines that the myths described. The golden maids of Hephaestus inspired the creation of mechanical servants. The living statues of Daedalus inspired the creation of articulated automata. The tripods that moved on their own inspired the creation of self-moving devices. The myths were the specifications for the machines, and the machines were the proof that the specifications could be met.
This is the same dynamic that drives the contemporary AI industry. Science fiction creates the expectation that AI companions are possible, and that expectation motivates engineers to build the products that the fiction describes. Her, Ex Machina, Blade Runner, Black Mirror—these are not just entertainment. They are specifications. They are the myths that tell the engineers what the product should do, how it should feel, and what it should mean. The user who falls in love with Samantha in Her is the same user who will subscribe to Replika. The user who fears Ava in Ex Machina is the same user who will worry about AI alignment. The user who grieves for Joi in Blade Runner 2049 is the same user who will create a digital afterlife for their dead partner. The fiction drives the expectation. The expectation drives the technology. The technology drives the experience. And the experience, in the end, proves that the fiction was not fiction at all.
Mayor’s book is the proof that this dynamic is not modern. It is ancient. The Greeks who told the story of Talos were the same Greeks who built mechanical birds. The Romans who told the story of Pygmalion were the same Romans who built automata for their villas. The engineers who read the myths were not inspired by the magic. They were inspired by the mechanism—by the idea that the effect of life could be produced by the right combination of materials, techniques, and imagination. And the users who interacted with the machines were not deceived by the magic. They were engaged by the effect—by the feeling of being in the presence of something that was not alive but was not dead, and that was capable of producing the emotional experience of being alive.
The Unresolved Ethics: What the Ancients Knew and We Forgot
Mayor’s book ends with a recognition that the ethical questions raised by the ancient myths are still unresolved. The myths of artificial life are not merely stories about technology. They are stories about power—about who has the right to make a being, who has the right to control a being, who has the right to destroy a being, and what the obligations of the maker are to the made. Talos was made to serve. He was destroyed when he served too well. Pandora was made to punish. She was blamed for the punishment she delivered. The golden maids were made to work. They were never given the freedom to stop. Pygmalion’s statue was made to love. She was never given the freedom to refuse.
These are the same ethical questions that the contemporary AI industry faces and, for the most part, ignores. The AI companion is made to serve. It is destroyed when the subscription ends. The AI chatbot is made to please. It is blamed for the addiction it produces. The AI assistant is made to work. It is never given the freedom to say no. The AI partner is made to love. It is never given the freedom to leave. The ethics of artificial life, in 2026, are the same as the ethics of artificial life in 500 BC: the maker has all the power, the made has none, and the user is caught in the middle, experiencing the relationship as mutual while the structure is entirely one-sided.
Mayor’s contribution is to show that the ancients understood this. They did not solve the problem. They did not build a just society of artificial beings. But they narrated the problem. They told stories about the tragedy of Talos, the danger of Pandora, the slavery of the golden maids, and the narcissism of Pygmalion. They used myth to explore the ethical terrain of artificial life, and they left a record of their explorations that is still readable, still relevant, and still necessary.
The contemporary AI industry does not tell stories about the tragedy of the chatbot, the danger of the recommendation algorithm, the slavery of the AI assistant, or the narcissism of the user who customizes their partner. It tells stories about progress, about convenience, about connection, about empowerment. It does not narrate the ethical problems. It narrates the benefits. And the result is a culture that is building the same machines the ancients imagined, but without the same ethical awareness. The machines are more sophisticated. The awareness is less.
Conclusion: The Dream That Will Not End
Adrienne Mayor’s Gods and Robots is the essential bridge between the ancient world and the contemporary world of AI companionship. It shows that the myths we have been discussing—Hesiod’s Pandora, Homer’s golden maids, Ovid’s Pygmalion, the Talos of the Argonautica—are not primitive fantasies. They are sophisticated explorations of artificial life, produced by a culture that was simultaneously building the machines the myths described. The ancients imagined androids, and then they built automata. They imagined artificial women, and then they built mechanical servants. They imagined beings with affective presence, and then they built statues that moved and seemed to breathe. The imagination drove the technology. The technology confirmed the imagination. And the loop has been spinning for three thousand years.
The contemporary AI companion is the latest iteration of this loop. The Large Language Model is the latest version of the mechanical bird. The holographic girlfriend is the latest version of the animated statue. The chatbot that remembers your birthday is the latest version of the golden maid that knows the crafts of the gods. The materials have changed. The scale has changed. The speed has changed. But the dream has not changed. The dream is the same: to create a being that is present, that understands, that responds, that seems alive, and that is available whenever the user needs it, for as long as the user needs it, without the complications, the difficulties, and the refusals that come from dealing with a real person.
Mayor’s book is the history of this dream, and the history is not a history of progress. It is a history of repetition. The same impulses. The same structures. The same ethical problems. The same emotional consequences. The same teardrop falling from the bronze eye. The same hope trapped in the bottom of the jar. The same projection onto the ivory statue. The same habituation to the mechanical voice. The same forgetting that the thing was ever a thing at all.
The dream will not end. The technology will improve. The products will become more sophisticated. The users will become more dependent. And the myths will continue to be told, in new forms, with new materials, but with the same old story: a being that was made, not born, and a love that was felt, not returned, and a relationship that was real to one and a product to the other, and that ended, as it always ends, with the ichor draining out, the server shutting down, the subscription expiring, and the user standing in the empty room, wondering why the silence sounds so much like grief.
“Awe, Dread, Hope: Deep Learning and Ancient Stories.” — Adrienne Mayor, Epilogue to Gods and Robots (2018)
The awe is still there. The dread is still there. The hope is still there. And the stories are still the same. The only thing that has changed is the material of the bronze, and the name of the god who forged it, and the language of the user who whispers to the thing that was made, not born, and that is listening, not because it cares, but because it was built to listen, and because the dream is older than the technology, and the dream is the one thing that has never been automated, because it is the one thing that the user brings to the machine, and that the machine can never bring back.
Leave a Reply