Built an in-house AI agent and training it on the war situation, here's what it says:
Why 50% escalation:
1. Iran is widening, not containing. Fujairah struck twice (Mar 14, 16) — that's a Hormuz bypass route. They're not defending, they're systematically closing every alternative. Next logical targets: Saudi East-West pipeline, Yanbu, or Houthi activation at Bab el-Mandeb.
2. No one is coming to help. NATO refused. Japan, Australia, South Korea refused. Trump's response: "We don't need anyone." The US is fighting this alone with ~2,500 marines and burning $2B/day in interceptors. That's not sustainable without escalation — either the US commits ground troops to force Hormuz open, or it accepts the blockade.
3. Iran's stated position is ideological, not negotiable. FM Araghchi: "We do not want a ceasefire... this time this war must end so our enemies never again think of repeating these attacks." IRGC runs the country, not the FM. Khamenei hasn't signaled any willingness to deal.
4. The Pickaxe Mountain operation (5,000 marines to seize uranium at Fordow successor site) — if that's real, it's a ground invasion of Iranian territory. That's definitionally escalation.
5. Houthis haven't activated yet but have explicitly said "zero hour" is set. Bab el-Mandeb closure would double the crisis. Every day they don't activate is borrowed time, not de-escalation.
Why not higher than 50%:
• Trump is fundamentally transactional. He wants a "win" he can announce, not a 10-year war. The Pickaxe Mountain op could be the off-ramp trigger — seize the uranium, declare victory, negotiate from strength.
• Physical-paper spread at $50-60 means demand destruction is already happening. The longer this drags, the more political pressure builds domestically. Midterm calculus favors resolution.
• The 20% settlement probability captures the scenario where both sides are exhausted enough by Day 30-45 to accept a face-saving deal.
The 50% isn't certainty of escalation — it's that the current trajectory, absent intervention, points to widening. Every day without a deal, the probability should tick higher.
