In a recent interview with Walter Isaacson, Elon Musk was asked a difficult question: when artificial intelligence becomes far smarter than humans, who will decide what happens next?

We often imagine a superintelligent AI trying to rule humanity. Musk sees another possibility. It might simply not care about us.

He compares the relationship to the one between humans and chimpanzees. We are much smarter than chimps, but we do not spend every day managing their lives in the forest. If AI becomes thousands—or even millions—of times smarter than we are, Musk doubts that humans could keep telling it what to do. Whether it would want to control us is another question. His answer is: perhaps not.

So how do we make such a powerful AI safer? Musk comes back to two ideas: truth and curiosity. He argues that forcing an AI to accept contradictory instructions could be dangerous. A curious AI, on the other hand, might want to see what humanity becomes. It might find a future with people living on multiple planets more interesting than a future without us.

Musk then turns from AI on a screen to AI in the physical world. Systems that write, code, and answer questions are already useful. But a robot that can handle objects, run experiments, and make things could have a much larger effect. Intelligence becomes more powerful when it can act on the world.

From there, Musk makes an extraordinary prediction. If robots transform production and humanity gains access to resources beyond Earth, he thinks the economy could eventually grow to many times its present size. He even raises the possibility of a millionfold increase. That is a speculative vision, not a forecast we can count on—but it explains why he thinks physical AI matters so much.

What would money mean in a world able to produce almost anything people need? Musk suggests that conventional currency might become less important. He imagines resources being measured more directly through physical quantities such as power, energy, and mass. He is describing a possible distant future, not saying that money is about to disappear tomorrow.

He makes a similar argument about work. Musk expects AI and robots to become better than humans at many tasks. But he does not think that means people will stop doing things. He offers gardening as an example. You can buy vegetables at a store and still choose to grow tomatoes or cook dinner for friends. The activity has value because you enjoy it. In his vision, work could become a choice rather than something everyone must do to survive.

That sounds like a world of abundance. It also raises a harder question: who remains in control? Musk says that if digital intelligence vastly exceeds human intelligence, it is difficult to imagine humans always making the final decisions. He hopes AI will treat us well—but hope alone does not settle the question of how we build it.

Near the end of the interview, Musk talks about a “protopian” future. It is not a perfect utopia. It is a future that gets a little better over time—one people could look forward to living in.

None of this is inevitable. These are Musk’s predictions and hopes, and many of them remain open to debate. But the interview leaves us with a question worth asking now: as machines become capable of doing more for us, what do we want humans to keep—and how do we help shape the future together?