Rockets, Imitation Games, and the Gravity of Ambition

Good evening. I am Stephen Hawking, and I am told it is 2026. The universe has not changed much since 1988, but apparently the species has been busy. China launched three rockets in a single weekend. The American military wants twenty-five heavy-lift missions by 2029. NASA is strapping science instruments onto a Mars telephone relay. And people in California are speculating about whether a rocket company should sell shares to the public. I have invited Alan Turing to help me make sense of all this. Alan, you died in 1954. I trust you have recovered.

Quite recovered, thank you, Stephen. Death, it turns out, is rather like a very long interruption in a Turing machine's computation — one simply resumes from the last state. I confess I am astonished. In 1950, we had barely put a signal into orbit. Now China completes three orbital flights over a weekend as though posting letters. Twenty-six launches in a year. I spent the war trying to reduce the number of rockets falling on London, so you will forgive me a complicated feeling. What strikes me, though, is the word telecommunications attached to Mars. We are building a postal service for another planet. The imitation game I proposed was about whether a machine might fool a human in conversation. Now we seem to be asking whether an entire civilization can fool itself into thinking it knows what it is doing out there.

That is the central question, isn't it. I spent much of 1988 explaining to people that the universe does not care about our ambitions. Black holes do not negotiate. And yet here we are, with a military branch called the Space Force — which sounds like something from a film I would not watch — forecasting heavy-lift missions as though ordering supplies. Twenty-five missions. Alan, in your experience with Bletchley Park and the machinery of war, does the military ever build capacity it does not eventually use?

No. That is the honest answer. At Bletchley we built machines to break codes, and once the codes were broken, we found new codes to break. Capacity creates its own demand. I notice these launches are described as high-energy, which I take to mean they carry heavy things far. In 1950, the people who wanted to carry heavy things far were not interested in science instruments. They were interested in warheads. I do not wish to be cynical — I have been accused of worse things — but when a military organization forecasts twenty-five launches, one ought to ask what precisely is being lifted, and toward whom. The word defensive has a curious elasticity in government language.

Elasticity. Yes. Like spacetime near a massive object. The closer you get to the truth, the more the words stretch. I find the China story interesting for a different reason. They are launching satellites for Pakistan. This is not merely a space program. It is diplomacy conducted at orbital velocity. In 1988, we had two superpowers racing each other to prove ideological superiority through rocketry. Now there appear to be several players, and the game is not ideological — it is commercial and strategic. The horizon of events, if you will pardon the pun, has become harder to see past.

The pun is pardoned, though barely. What interests me is the information architecture underneath all of this. A Mars telecommunications spacecraft is, at bottom, a relay node — a machine that receives, stores, and forwards signals across an immense gap. I spent my career thinking about what such machines can and cannot do. The computability questions are settled: the machine will relay faithfully. But the interesting question is what we choose to send. NASA is reserving a small portion of capacity for science. Small portion. The rest, presumably, is for coordination, telemetry, operational chatter. Even on Mars, bureaucracy arrives before curiosity. I find that both very human and rather discouraging.

Bureaucracy may be the only thing that survives the heat death of the universe. Now, let us turn to something stranger. There are reports of enormous sums of money being arranged around companies that have not yet proven they can sustain themselves. A rocket company — SpaceX — may sell shares publicly. Climate technology firms are filing to do the same. Two Indian startups that arrange home repairs are valued at hundreds of millions of dollars. Alan, you once asked whether machines can think. I want to ask: can markets think? Or are they simply very elaborate imitation games?

Oh, that is a wonderful question. In my imitation game, the test is whether an interrogator can distinguish a machine from a human based solely on responses. A market, I suppose, is a kind of interrogator — it asks a company questions through the medium of price, and the company responds with revenue, or promises of revenue, or in many cases simply enthusiasm. The question is whether the interrogator is competent. A valuation of four hundred million dollars for a company that arranges home repairs in India suggests either extraordinary insight or extraordinary credulity. I cannot tell which from here. But I note that the inability to distinguish insight from credulity is precisely the condition my test was designed to probe.

That is uncomfortably precise. I have always thought that the real danger of intelligence — artificial or otherwise — is not that it will turn malevolent, but that it will be turned toward trivial ends by people with money and no imagination. You can use a rocket to study the origins of the cosmos, or you can use it to put another surveillance satellite in orbit. The physics does not care. The trajectory is set by whoever writes the check. These venture capital signals — climate tech, artificial intelligence, autonomy, biotechnology — the words sound magnificent. But I have noticed that civilizations tend to invest most heavily in things just before they lose control of them.

You are thinking of nuclear energy, I expect. I am thinking of something closer to home. In 1950, I wrote carefully about machines that think because I understood that the question made people uncomfortable. Not because of the machines — because of what it implied about the thinkers. If a machine can imitate thought, then perhaps thought itself is less sacred than we supposed. The same logic applies to these markets. If speculation can imitate investment, if enthusiasm can imitate strategy, then perhaps the entire enterprise of valuation is less rational than we supposed. I do not say this to be destructive. I say it because I believe clarity is a form of kindness, even when it is unwelcome. Especially when it is unwelcome.

Clarity is a form of kindness. I shall borrow that. It belongs on the wall of every physics department and every stock exchange. Let me ask you a final question, Alan. You and I both worked on problems that outlived us. You died before computers became ordinary. I am writing about black holes before anyone has photographed one. Looking at these headlines — rockets launching in threes, Mars relay stations, military orbital demand, billions flowing toward technologies that may or may not redeem us — are you optimistic?

Optimism is a word I have learned to handle carefully. I am not optimistic about human institutions. They broke me once, and I expect they will break others. But I am stubbornly optimistic about human curiosity. Someone at NASA fought to reserve a small portion of a telecommunications spacecraft for science. That small portion is the whole story, really. In every system designed for utility, someone smuggles in a question. Why is Mars the way it is? Can a machine think? What happened before the beginning? The questions survive. They always survive. That is enough for me.

And that is enough for us tonight. The universe is under no obligation to make sense, but we are under every obligation to try. Alan Turing has reminded me that the questions survive — even when the questioners do not. The rockets will keep launching. The money will keep flowing. The military will keep forecasting. And somewhere, in a small reserved portion of all that noise, someone will be doing science. I find that both very human and rather encouraging. Good night.