How this works

This is an AI model of me, built from my own writing and a personality model. It answers in my voice. It is not me live.

It works from material I have supplied about what I have said, done, and believe. It selects supporting passages from that material as it forms an answer. The aim is to separate what is known from what is inferred, and say when something is missing. It can still get things wrong.

When one of my recordings directly addresses your question, it appears above the written answer. A passage starts at a published chapter, or you can play the full video. You always have the original recording and its date. The text is a separate second-brain answer, not a transcript of what I said.

Playback loads YouTube when you press play. If the embedded player is unavailable, the YouTube link opens the recording at the same starting point.

It will not go anywhere that would harm another person, including the specific people in my life. If you surface a blind spot I have not named, it logs the observation and thanks you, and I review it later.