Jeremiah Chronister

I run AI systems in a regulated domain, where a mistake is expensive and the feedback arrives slowly. Before that I ran quantitative models on live capital and operated a registered fund, where the feedback arrives immediately and the mistake is priced the same day.

The through line is the same discipline in both places. Position over opinion. Exposure over belief. Structure over mood. The fine print over the loud story, and verification against the primary source rather than against a confident summary.

I write every two weeks about what that discipline looks like in practice: operating AI where the consequences are real, thinking through a capital decision structurally, and the long horizon that the rest of it is for.