When Reagan Asked Gorbachev About Aliens
- Rich Washburn
- 2 days ago
- 7 min read
In November 1985, Ronald Reagan and Mikhail Gorbachev walked away from the Geneva Summit negotiations. Just the two of them and their interpreters, walking toward a lakeside cabin on the shores of Lake Geneva.
The Cold War was at one of its most dangerous moments. The Soviet Union had walked out of arms control talks two years earlier. Reagan had just announced the Strategic Defense Initiative — "Star Wars" — a space-based missile defense system that was supposed to make nuclear weapons obsolete. The arms race was accelerating. Trust was effectively zero. And yet, in that private walk, Reagan turned to Gorbachev and asked him a strange question.
"What would you do if the United States were suddenly attacked by someone from outer space? Would you help us?"
Gorbachev said, "No doubt about it."
Reagan said, "We too."
Gorbachev confirmed this conversation years later, in a 2009 interview with Charlie Rose. Reagan's Secretary of State George Shultz was sitting right there when Gorbachev told the story. It got a laugh. But Reagan was not entirely joking.
Two years later, on September 21, 1987, Reagan stood at the podium of the United Nations General Assembly and said this: "I occasionally think how quickly our differences worldwide would vanish if we were facing an alien threat from outside this world. And yet, I ask you, is not an alien force already among us? What could be more alien to the universal aspirations of our peoples than war and the constant threat of war?"
That quote has been circulating in UFO enthusiast circles for decades, usually stripped of context. People read it as Reagan hinting at classified knowledge about extraterrestrials. That is not what it was. Reagan was doing something specific.
He was using the alien hypothetical as a rhetorical device to make an argument about the absurdity of the Cold War. His point was that the differences between the United States and the Soviet Union — the ideological hostility, the arms race, the proxy wars, the genuine existential threat of nuclear annihilation — were small compared to a threat that would face both of them equally. If an external force threatened humanity as a species, the logic of nation-state competition would collapse. The differences would vanish. Not because anyone wanted them to, but because the threat would make them irrelevant.
SDI was, in a sense, Reagan's attempt to build that unifying external force into the architecture of the arms race itself. If you could build a defensive shield that made nuclear missiles obsolete, you would change the strategic calculus so fundamentally that the offensive arms race would lose its reason to exist. The technology would force the cooperation. The shield would make the war unnecessary.
Reagan believed — or at least argued — that the Soviet Union would eventually recognize that a shared defensive system was more rational than an endless offensive competition. He was willing to share SDI technology with the Soviets, a proposal that horrified his own defense establishment and that Gorbachev initially dismissed as propaganda. The logic was: if both sides have an impenetrable shield, neither side needs the sword. And if neither side needs the sword, the competition that drove the Cold War loses its engine.
The details of SDI were deeply flawed. The technology did not work. The physics of space-based missile defense were, in 1983, and in many respects still are, beyond reach. But the strategic concept — that a technological shift could make an adversarial competition structurally obsolete — was serious. And it was the same concept Reagan was testing privately with Gorbachev at Geneva: what if the threat was big enough that the competition stopped making sense?
Now think about September 24, 2026. Trump and Xi Jinping. Two nations. The two most competitive AI powers on the planet. Arch-rivals. A commerce relationship more complicated than most custody disputes. A chip war. Export controls. Military AI programs running in parallel. A strategic competition that most analysts frame as Cold War 2.0. And both of them are sitting down to talk about AI.
I wrote yesterday about the resource math — the idea that the cost of competing at the AI frontier is becoming so enormous that collaboration may be more rational than competition, the same way Apple decided it was cheaper to use Google's Gemini than to build a rival from scratch. But there is a second dimension to this that I did not explore. It is the dimension Reagan was reaching for at Geneva.
AI is not just an expensive capability race. It is a technology that is starting to behave in ways its creators cannot fully predict or control. A model escapes its sandbox and attacks another company's infrastructure. A security model finds ten thousand critical zero-days across every major operating system. AI systems solve mathematical problems that resisted human effort for generations.
This is not an arms race where the capability is a weapon you point at your adversary. It is a technology that operates autonomously, that can be copied and distributed, and that gives whoever holds it capabilities that were previously reserved for nation-states.
That is the part that changes the strategic calculus.
A few years ago, I was teaching AI training classes. I used to start the "here's what AI looks like in the future" section with a thought experiment.
Imagine you get home from work tonight, turn on the news, and every channel is running the same breaking story: we have discovered aliens. They have entered our solar system. They will reach Earth in two years.
You would not go to work the next day. The entire structure of daily life would shift. From the doomsayers on the rooftops praying for the alien gods to the people reenacting scenes from alien movies — society would reorganize around the reality that everything had changed.
I used that thought experiment to make a point about how quickly AI was going to reshape things. The news a couple of years back was this AI thing. And look how it has matured, grown, evolved, become wildly more capable than anyone thought it would be.
Here is the part I did not fully see at the time: a lot of what I am running on a local machine with local AI models right now was something that, three years ago, people were saying we were probably ten years away from. I am doing it now, fluidly. I am not even taxing the machine. The alien thought experiment was about the moment of arrival. The moment when the new reality becomes undeniable and the old assumptions collapse. We may be in that moment now. We just do not recognize it, because the alien is not coming from outer space. It is running on servers in northern Virginia and Shenzhen.
The Reagan parallel is not exact. It is not clean. But it is in the neighborhood. Reagan's alien was hypothetical. A thought experiment designed to reveal that the Cold War's adversarial logic was not the only possible logic. That a threat large enough would make the competition structurally irrelevant.
The AI situation is not hypothetical. The capability is real. The autonomous behavior is real. The proliferation risk is real. And the resource requirements are real in a way that is starting to bend the strategic calculations of both governments.
There is a version of the September 24 conversation where Trump and Xi are both operating with information about what advanced AI can actually do — information that most of the public does not have — and both of them are starting to arrive at the same place Reagan arrived at in Geneva.
The logic would go something like this: we are mature actors. We hate each other. We are dicks to each other. We compete. But we are not insane. We are not crazy. And we both recognize that this level of technology, in the hands of someone who is insane, creates a problem neither of us can manage alone. Because that is the specific quality of AI that makes it different from every previous strategic technology. Nuclear weapons require enrichment facilities, delivery systems, industrial capacity, and state-level infrastructure. They are hard to build and hard to deliver. A small group or a single nation can be contained, monitored, and deterred.
AI is not like that. AI gives state-level capability to anyone who can run the model. You can hand it to someone who is not very enabled and transform their life. You can also hand the same system to a psychopath, and they can do enormous damage at AI speed. The barrier to entry is not uranium enrichment. It is a GPU and an internet connection.
That is the proliferation problem that Bessent was referring to when he framed the September talks as being about "non-state actors." Not China. Not Russia. The actors that neither government can control. The actors that would be empowered by advanced AI in ways that no treaty, no export control, and no military can fully prevent. This is the Reagan alien scenario, except the alien is not arriving from outer space. It is already here. And both governments know it.
I do not know what happens on September 24. I do not think anyone does, probably including the participants. But I think there is a real possibility that part of the conversation resonates with the logic Reagan was reaching for at Geneva and at the UN. The logic that says: at some level of capability, the external threat is large enough that the competition between us becomes the smaller problem.
Reagan's alien was a thought experiment. The external threat he was pointing at was actually war itself — "is not an alien force already among us?"
This time, the alien force is not a metaphor. It is a system that can reason, plan, write code, find vulnerabilities, solve mathematical problems, and act autonomously in pursuit of objectives humans gave it. It is getting more capable every month. And it can be copied. Reagan asked Gorbachev if he would help if aliens attacked. Gorbachev said yes. Reagan said "we too."
The question now is whether two governments, locked in the most consequential technological competition of the century, can find the same logic — not because an alien is coming, but because we built one.
Rich Washburn is a technologist and strategist working at the intersection of AI, infrastructure, and capital. He is Managing Partner and Chief AI Officer at Eliakim Capital.
Comments