Field Notes on Trust and AI

RadiationBox

An independent publication on trust signals, media incentives, and AI-native behavior.

Latest Published Essay

Part 7: How I Built The Analysis Lane Behind This Project

I wanted to end the series with the less glamorous part: how the analysis lane worked in practice. The pattern that held up was plain enough: pull public posts only when the question justified it, keep the result in one authority, categorize it carefully, and do the later reading from that stored record instead of buying the same lesson twice.

continue reading

Also Published

Part 6: What Looking Back With X Taught Me About A Campaign I Already Knew

The first five pieces in this package do the modular work. They separate why the X API still mattered to me , what changed from the old Twitter API to the X API now , how the He Gets Us dataset got built , what the surface findings look like , and what the deeper read adds .

continue reading

Also Published

Part 5: The Deep Read: Attention, Narrative Tremors, And What The Math Adds

The surface read already tells you this campaign creates argument. The deeper read tells you what kind of argument matters most.

continue reading

Also Published

Part 4: The Surface Findings From 50,000 He Gets Us Posts

The simplest honest summary of the He Gets Us X dataset is this:

continue reading