We spent twenty years learning why people stay unknown. Then the constraint disappeared.

I’ve spent more than twenty years inside the machinery organizations use to understand people — advising boards, university presidents, and senior leaders, building models of individual capacity, and watching what those models could and couldn’t do. Two observations wouldn’t leave me alone.

The first: the decline is quiet, and it happens between the touchpoints. The session ends, the review cycle closes, the cohort graduates — and each person returns to the long, unobserved stretches where their actual life happens. That’s where momentum stalls, where the wrong turn gets taken and no one notices for two years. Everyone in development knows this; almost nothing is built for it, because nothing could be. Continuous, individual attention didn’t scale.

The second was personal. Years ago I built a firm around deep, individualized capacity work — the kind where you come to know a person before presuming to advise them. It worked. And at the moment it mattered, the verdict from one of the most sophisticated institutions in the world was simple: you can’t scale it. They were right. So the world kept doing what it could afford — training to groups, measuring to the average — and every individual remained, in the ways that matter most, unknown.

What changed with AI isn’t that machines got clever. It’s that sustained, individualized attention became possible for the first time — attention that remembers, integrates multiple ways of knowing a person, and compounds over years instead of expiring at the end of a session. The thing that couldn’t scale, scales.

Amber is what we’re building with that possibility: a longitudinal individual intelligence that begins with rigorous science, learns each person’s story, and stays. Not another tool that hands you a score and disappears. A relationship that helps your life make sense over time.