Here's the blueprint: https://hren.io/files/Meditation.pdf
The AI would need to read the emotion through brainwaves (unless there’s a better way) and apply changes to it via controller variables (speed, size, movement, colors,…) and based on instant feedback “hold” the user in that "zone".
Can this be created with desired results?
I see a lot of use cases and benefits from it but think it should be "open source" and not for commercial use. I see a lot of bad things and manipulation you could do with it as well so some control would be needed.
The visuals can be many different experiments. I've found that Victor Vergara's Perlin noise could be quite interesting as the AI can control a lot of its variables: https://codepen.io/vcomics/full/djqNrm/
Anyone wants to help with the prototype and try to create this?
The tech stack I thought for the proto would be React/React Native, Tensorflow JS, three.JS. But I'm missing some parts and technical knowledge to pull this off.