What this covers
Annie Jacobsen and Tyler Cowen debate the rationality and necessity of nuclear weapons in a conversation that pivots on two opposed views of how to prevent catastrophe. Jacobsen argues that nuclear war is fundamentally unwinnable—that game theory applied to it breaks down because no rational actor would kill hundreds of millions to avoid losing first—and that the only viable path to safety is disarmament and direct communication between powers. Cowen counters that deterrence is the least-bad option available, that fewer weapons are not obviously safer, and that the alternative to nuclear weapons may be worse. The conversation draws on declassified war games, technical analyses of missile defense, historical precedent, and the psychology of risk and belief.
The specific terrain they cross is wide. On the technical side, Jacobsen establishes that missile defense cannot reliably intercept ICBMs—a single interceptor has less than 50 percent success even in favorable test conditions—whereas Cowen argues the US should spend 4-5 percent of GDP on defensive systems. On policy, Jacobsen points to Ronald Reagan's reversal toward disarmament after watching The Day After and his subsequent coordination with Gorbachev as proof that communication works; Cowen warns that reducing arsenals may eliminate second-strike capability, which he sees as stabilizing. They also collide on the six-minute window for launch decisions—Jacobsen calls it irrational and insists no single person should hold that power; Cowen treats it as an acceptable cost of the deterrent system. A third thread examines conspiracy thinking, the placebo effect, and claims of ESP—material Jacobsen drew from her reporting on classified military programs, where she argues the Defense Department pursued paranormal research that couldn't be operationalized.
Jacobsen argues that nuclear war is fundamentally unwinnable madness in which any US involvement leads to total annihilation, so the only viable path to safety is disarmament and communication rather than missile defense or larger arsenals; Cowen presses the contrary view that deterrence is the least-bad form of madness and that fewer weapons are not obviously safer.
- War games like Proud Prophet show that however nuclear war begins, if the US is involved it ends in Armageddon
- Missile defense cannot reliably intercept ICBMs, so building defensive systems is a false hope
- Reagan's reversal toward disarmament after watching The Day After is the only demonstrated pathway out of the madness
Nuclear weapons deter war through their perceived madness, which is a rational basis for keeping them.
- Nuclear war may be insane, but it could be the better insanity among available options; if enough people regard it as madness, it won't happen, which is fine reason to keep nuclear weapons.
“It is insane, but it might be the better insanity of the ones available to us.”
- People are overly inclined to believe in conspiracy theories because, just as humans once attributed weather to gods, it is psychologically easier to think of disliked actors as plotting against us than to accept that events are a series of accidents.
“Just as we used to attribute the weather to gods, we anthropomorphize many things. It’s easier to think in terms of the people we don’t like as plotting against us rather than things being a series of accidents.”
Second-strike capability deters war more effectively than disarmament, which can destabilize deterrence.
- As nations move toward disarmament, some may lose second-strike capability, which can make the situation more dangerous; second-strike capability (as Israel now has) is stabilizing.
“don’t either we — or some other countries in that process — lose second-strike capability? And that makes the situation more dangerous. I’m very glad Israel has second-strike capability now, and they didn’t a few decades ago.”
Nuclear weapons are preferable to the biological, chemical, or AI weapons that would replace them.
- The alternative to nuclear weapons is not peace but other deterrents like biological, chemical, or AI weapons; since we have 75+ years of experience managing nuclear weapons, switching to mutual assured destruction via biological weapons would not be safer.
“the alternative to nuclear weapons — it’s not sticks and stones or pistols. It could be biological weapons, chemical weapons, many other nasty things. It could be AI as a kind of weapon.”
The US should invest heavily in defensive systems to protect major cities from nuclear attack.
- If nuclear war is as unthinkable and destructive as claimed, the US could and should spend around 4-5 percent of GDP to build a defensive system protecting at least Washington and New York, just as the Manhattan Project consumed a large chunk of GDP.
“if we spent 5 percent of GDP, we could come up with something. It wouldn’t be perfect, but it could protect Washington and New York, right?”