On: Deadline for US Congress on war in Iran

23 June 1954

The news reports speak of “plans” and “options” - as if war were a problem to be solved by computation. But before one asks how to strike Iran, one must ask whether the question itself is computable. Can “deterrence” be formalized into a decision procedure? Does there exist any finite sequence of military actions guaranteed to produce a stable outcome?

The generals speak of “short and powerful” strikes, as if violence were an algorithm with predictable termination. But war is not a Turing machine. There is no halting state where the opponent simply stops. Each action begets unpredictable reactions, branching into an infinite tape of consequences. The problem is undecidable - not for lack of intelligence or precision, but because the system itself is non-computable.

They imagine they are optimizing a solution. In truth, they are iterating a function with no fixed point. The real question is not “how to strike,” but “can the desired outcome be computed at all?” If not, then no amount of force will converge on stability.

I see the same error everywhere - in diplomacy, in policy, in the crude mechanical thinking that mistakes power for control. The map is not the territory. The procedure does not halt.