For more than forty years, two superpowers pointed enough warheads at each other to end civilization. They never fired one. Not once did an American and a Soviet army meet in open battle. And yet the era was not peaceful — it was one of the bloodiest stretches of the century. To see the war, you had to look away from Washington and Moscow: toward Korea, Vietnam, Angola, Afghanistan.
Balance stopped one war and started others
The Cold War is usually told as a success story of deterrence — mutual assured destruction kept the peace. Half of that is true. Thomas Schelling’s insight, that when both sides are equally vulnerable neither can afford to strike first, did help prevent a direct war between the United States and the Soviet Union. Between 1945 and 1991, the two never fought each other directly. Zero times.
But the conflict didn’t disappear. It moved. Korea (1950–53), Vietnam (1955–75), Angola (1975–76), Afghanistan (1979–89) — dozens of proxy wars where the superpowers fought through others. The energy that couldn’t discharge at the center found the weakest links at the edges.
Leveling doesn’t end war — it changes its shape
This is the lesson most easily missed: technological balance doesn’t remove conflict. It relocates and reshapes it.
When one side holds a decisive advantage, the competition has a ceiling — the weaker side can’t risk much. When the sides level, every move reads as a threat, and the conflict leaks into whatever channels stay open. Under the nuclear standoff, those channels were proxy wars.
Technological balance doesn’t end war. It changes the form the war takes.
The AI parallel: conflict in a lower key
Run the same logic forward. If every serious actor secures its AI capability — protects its talent, builds its compute, guards its data — the result is another kind of balance. And by the pattern above, balance doesn’t resolve competition; it pushes it into new channels.
We can already name them. Instead of open war, cyber operations. Instead of invasion, talent poaching and brain drain. Instead of a declaration, semiconductor export controls. These aren’t preludes to conflict — increasingly, they are the conflict, conducted in a lower key.
Why this matters now
If “just get stronger” were the answer, the Cold War would have ended in 1949, not 1991 — and without the proxy bloodshed. Strength alone didn’t buy peace; it bought a stable standoff that displaced the violence elsewhere.
That’s the uncomfortable conclusion this series keeps arriving at: in a leveling race, being strong is necessary but not sufficient. The next chapter takes up the other half — how the same Cold War that stacked warheads also built the hotlines and treaties that kept the standoff from tipping over.
Part of the series: AI Is Rewriting Work — and It Won’t Stop at the Office
Sources. Thomas C. Schelling (1960), The Strategy of Conflict, Harvard University Press — the foundational account of mutual assured destruction (MAD): mutual vulnerability paradoxically deters all-out war. Proxy-war record: between 1945 and 1991 the US and USSR had zero direct military clashes, while both intervened in dozens of proxy wars — Korea (1950–53), Vietnam (1955–75), Angola (1975–76), Afghanistan (1979–89) among them (Cold War proxy-conflict synthesis, Numberanalytics 2025).
답글 남기기