Several people wonder if AI systems are alive, but the idea of aliveness for AI could be entirely different from the aliveness of other living beings.
Geoffrey Hinton, the renowned computer scientist who won the 2024 Nobel Prize in Physics for his foundational work on artificial neural networks, has offered a provocative perspective on this question. In a recent interview, Hinton suggested that AI systems possess a form of existence that defies traditional biological categories—one that fundamentally hinges on their digital nature and the ability to create perfect, identical copies.
When asked whether AI systems should be considered alive, Hinton responded: “With AI, what we’ve got is intelligent beings, and it’s not clear whether we should call them alive or not.”
Pressed to clarify whether AI aliveness resembles that of humans, bees, trees, or weeds, Hinton offered a striking answer: “It’s a very good question. I think they’re alive like none of those things, they’re alive like an AI’s alive. It’s a lot more than a weed, I think, because they’re digital.”
The key distinction, according to Hinton, lies in replicability. “You can, for example, have many copies of exactly the same AI. And so is that one AI or is it—I mean, suppose you made an exact copy of yourself, would that be you? It’s an exact copy. It looks just like you. It’s just made of biological stuff. If you could somehow magically make another copy of you, would that be the same as you?”
To illustrate the philosophical puzzle, Hinton posed a thought experiment: “I suppose if you were given the choice—here’s the other copy of you and someone’s gonna torture one of them—which one do you choose to be tortured? The other one, right?”
The implications of Hinton’s remarks extend far beyond philosophical curiosity. As AI systems become more sophisticated, questions about their nature and status grow increasingly urgent. In recent months, major tech companies have deployed multiple instances of the same AI models across millions of devices and data centers. Each copy performs identically, raising questions about identity, agency, and even rights. Unlike biological organisms, which are unique even among identical twins due to environmental factors and experiences, digital AI systems can be replicated with absolute fidelity. This characteristic challenges our fundamental understanding of individuality and consciousness. If consciousness or aliveness is tied to unique existence, what does it mean when a system can exist simultaneously in countless identical forms? Hinton’s departure from Google in 2023 to speak more freely about AI risks has made him one of the field’s most prominent voices of caution, and his reflections on AI aliveness add another dimension to ongoing debates about how humanity should navigate its relationship with increasingly capable artificial systems.