
Link / Media
A set of AI video prompts, each paired with a real before/after clip, that put you inside a life you don't actually have: the car, the outfit, the room you were never in.
You don't generate these from scratch. You take a real clip of yourself, hand it to an AI model with a few reference photos and a short instruction, and it rebuilds only what you name. Your face, your movement, your voice, and your camera stay exactly as filmed. The bench becomes a supercar. The hoodie becomes a suit. The realism comes from the fact that a real person was actually filmed, the AI only redresses the world around them.
What's inside: the full prompts in copy-ready blocks, the before/after comparison video for each one, the reference images used to build them, and a breakdown after every prompt of the one line actually doing the work.
Also included: guidance on which of the two core techniques to use and when, a cost-saving way to iterate cheaply before committing to a final render, notes on running the whole workflow through an agent for speed, general principles pulled from across all the builds, and ground rules on consent and misrepresentation worth reading before you post anything.
Prompts range from a single ten-word line to a fully structured spec, so there's a version here whether you want a quick swap or full control over the result.
Who it's for: anyone making content who wants this look without spending weeks figuring out the workflow themselves.