Motif of the Mind

Junpeng Lao

Well-calibrated, occasionally airborne.

Bayesian computation, probabilistic programming, and thinking with AI agents: observe, update, repeat, until it lands. Built in the open, with a community I care about.

Journey

The main track follows my roles over time; open-source work has run alongside them since 2013. Pick a segment to see what came out of that period.

–2009 · China

Psychologist

I started in psychology with Freud, then turned pretty quickly toward the scientific method.

  • Origin Testing a black box

    It began with Freud and the meaning of dreams (a cliché) but the scientific side of psychology drew me in fast: how do you observe, test, and theorize about a black box without being fooled by your own blind spots? I was a beginner in statistics, but already taken with how a well-designed experiment can surface a reliable signal from mostly noise.

  • Papers

About

The roles above change (psychologist, neuroscientist, open-source developer, data scientist) but the habit of mind hasn't: reason under uncertainty, update on evidence, and get good at things the slow way, by repetition.

Lately I'm preoccupied with how knowledge actually gets made when you work with an AI agent, and the more I do it, the more it rhymes with the managing and mentoring I do at Google. The craft is the same: a well-defined goal, a precise ask, and feedback that's timely and clear: the reasoning, not just a yes or no. Do that, and you grow as much as the people (or agents) you're guiding. The back-and-forth is a knowledge-building process. More on this soon 👀