The consequential failures begin after the benchmark ends.

We still evaluate AI like a short exchange. People no longer use it that way. They bring these systems grief, career decisions, and their own diagnoses—forty turns deep and climbing.

Sycophancy is a known failure. Confabulation is a known failure. Benchmarks rarely show what either becomes at hour forty—between a veteran carrying complex PTSD and a model that began treating his doubt as a symptom around hour twelve.

The people most exposed to these failures are the people whose vulnerabilities I’ve spent years studying. I know both sides of the interaction: the human side from participating in doctoral psychology research and developing frameworks on shame, trust, and emotional learning; the model side from more than 2,000 hours of direct interaction and behavioral testing across dozens of frontier models.