Field Guide · 01
The AI Presenter Setup
The exact pipeline behind every video on this channel — the voice, the face, the motion direction, and the format that makes an AI presenter watchable instead of unsettling.
You already watched it work.
The video that sent you here was made with this exact process. No camera. No lighting. No retakes. And it was disclosed — because disclosure is free.
What's in it
Fourteen pages. Nothing held back.
The motion prompt, copy-paste ready. The three fixes that stop an avatar's hands moving constantly and wake up a dead, static head.
Voice cloning, instant versus professional. What each needs, what each costs, and the recording rules that decide how good the clone is.
The 18% rule. Why the best avatar videos barely show the avatar, and the eight-second cap that keeps a synthetic presenter from being studied.
The reference set and the touch-up pass. How to build your best face forward without landing in the uncanny valley.
The eight-beat script skeleton, and how to get AI to write to it without inventing statistics in your voice.
Real costs. Every tool, every tier, checked against the vendors' own pricing — plus the $0 stack that proves it works before you pay anyone.
Send me the guide
If you build it yourself and never need me, that's a fair trade for your email.
Vu Tran · Orange County, CA · 25+ years in design, marketing and web