Most AI research studies the machine. We study the interaction.

Ivanooo runs a research program on human–AI interaction: how people lose command of AI without noticing, and what it does to their work, skills, and brands. About 79% of AI failures leave no visible trace; ordinary monitoring catches ~12%. People adopt AI output without engaging their reasoning, and confidence rises even when the AI is wrong.

We never ask — we read the work. Real transcripts, the episode as the unit, signals before labels, nothing published below Cohen's κ ≥ 0.70. The research produced the AI Operator framework — Fluency × Direction, the Four Moves, Cognitive Surrender, the Darkness Library — plus an original failure archetype: silent tool failure.