Overview of Why Nuclear War Is More Likely Than You Think (with Carlo Rovelli)
This episode of The Rest is Politics launches a new series on existential risk with physicist and author Carlo Rovelli, focusing on why nuclear war may be more likely now than many people assume. Rovelli argues that the danger has not disappeared with the end of the Cold War—in some ways, it has grown. He points to changing geopolitical dynamics, the rise of artificial intelligence in military decision-making, and a renewed push for rearmament in Europe as major reasons for concern.
Why Rovelli Thinks the Nuclear Risk Has Increased
Rovelli’s core argument is that nuclear war feels “unfashionable” precisely when it should be taken most seriously. He says public attention has faded since the 1960s, which creates complacency.
Main factors he identifies
-
Public forgetting and political neglect
Nuclear risk is no longer central in political debate, especially in Europe, even as rearmament discussions intensify. -
A multipolar nuclear world
The Cold War’s two-superpower structure has given way to a more complex landscape with the U.S., Russia, and China as major nuclear powers. -
Artificial intelligence in military command systems
Rovelli warns that AI can speed up decision-making in a crisis, reducing the time humans have to intervene and increasing the chance of catastrophic error. -
Rising tensions and arms buildup
Rearmament, especially in Europe, can increase instability by encouraging fear, escalation, and counter-rearmament by rivals.
Nuclear Deterrence, Fear, and Who Is Most Likely to Use Weapons
A major part of the discussion centers on which states are actually most likely to use nuclear weapons.
Rovelli’s view on “small” nuclear states
He argues that smaller nuclear powers like North Korea are less likely to initiate a strike because doing so would almost certainly mean their own destruction. In his view:
- Mutually assured destruction works best when both sides are clearly vulnerable
- Smaller states tend to use nuclear weapons primarily as deterrence, not as offensive tools
- The real danger arises when a state believes it faces an existential threat and the other side is not fully nuclear-armed
The most dangerous scenario
Rovelli sees the biggest risk not in symmetrical nuclear standoffs, but in situations where:
- one state believes it is facing annihilation,
- it has nuclear capability,
- and it thinks a preemptive strike may be the only option.
He suggests that this is why cases like Israel/Iran, India/Pakistan, and North Korea/South Korea are especially dangerous in principle, because fear and perceived survival threats can overwhelm rational restraint.
AI and the “Kill Chain” Problem
The conversation digs into the dangers of integrating AI into nuclear command systems.
Key concern
AI may shorten the “kill chain” — the sequence of decisions between detecting a threat and launching a response. That creates pressure to delegate more authority to machines.
Why Rovelli worries
- AI is fast, but not fully reliable
- It can hallucinate or behave rigidly
- It may push leaders toward overly technical, formulaic decisions
- Governments face a trap: not using AI could leave them at a strategic disadvantage, but using it may make war more likely
Europe, Rearmament, and the Risk of Escalation
A substantial portion of the discussion focuses on Europe’s response to Russia and the push to increase military spending.
Rovelli’s warning
He argues that European rearmament may:
- increase Russian paranoia,
- deepen tensions,
- and create a self-fulfilling spiral toward conflict.
He is especially critical of the idea that spending heavily on nuclear systems, such as Britain’s Trident program, strengthens security without increasing overall danger.
His broader political point
Rovelli suggests that Europe’s long-term interest is not endless escalation, but eventually finding a way to live with Russia again. He frames the alternative as a dangerous repeat of historical European self-destruction.
Historical Warnings and the Danger of Complacency
To underline how disasters happen when people assume they won’t, Rovelli and the hosts invoke historical examples:
- the collapse of Mussolini’s Salò Republic, where officials were still debating routine reforms shortly before collapse,
- the outbreak of World War I, which few expected right before Sarajevo.
The lesson: major catastrophes often arrive when elites believe they have managed to avoid them.
Final Takeaway
Rovelli’s message is that nuclear danger has not gone away — it has become more complex and, in some ways, more dangerous. His main warning is that:
- complacency is itself a risk
- AI may make nuclear crises harder to control
- arms races can increase, rather than reduce, the chance of catastrophe
- the world should talk about nuclear war more, not less
The episode ends by directing listeners to sign up for the full interview at therestispolitics.com.