For most of the past two years, the scarce input for AI was chips. Increasingly it is electricity. Companies can order GPUs, but they cannot easily get a few hundred megawatts of grid power delivered to a site, and the wait for a new grid connection in many US regions now runs for years. That is why data center deals are announced in gigawatts, why tech companies are signing contracts to restart nuclear plants, and why some sites are building their own gas turbines.

Backcasting is a planning method that starts with a defined future outcome and works back to the present, asking what has to happen (and in what order) for that outcome to be reached. Forecasting runs the other way. It takes today’s trends and projects them forward to see where they lead. That sounds like a small difference in direction, yet it changes the question completely. A forecast asks what is likely.

Every escalation cycle in the Gulf eventually runs into the same hard constraints: oil flow, market stability, and military sustainment. When pressure builds too fast, something has to give—not because leaders suddenly prefer diplomacy, but because systems start to strain. A ceasefire, especially one announced loudly and ambiguously, acts as a release valve. The constraints are structural, not psychological. Oil markets respond to threat premiums with price spikes that damage the economies of both belligerents and bystanders alike.

First, one important correction matters. The Strait of Hormuz has not been shut in a neat, absolute sense. Traffic can fall sharply, access can become selective, insurers can pull back, and naval presence can reshape behavior long before a formal “closure” exists. That distinction sounds technical, but it changes the analysis. Markets react not only to blocked geography, but to uncertainty, risk pricing, and the creeping sense that passage is no longer neutral.