Moonshot Tesla AI Idea

Tesla is detecting the world by detecting objects in the world and applying rules to them.

For each object in a scene Tesla will decipher what that object is. If it's not road, then...

How cool would it be if Tesla was using FSD computers to detect these things and apply different sets of rules.

One popular application right now is Robots. Apply FSD to robots makes good sense.

But another aspect Tesla is working on is scene building. I think I thought of a clever application that may make an appearance sooner rather than later.

Being in LA for so long, Elon was surrounded by comedians, movie stars, and many different backgrounds in anything film.

Enter the Tesla FSD computer into Hollywood. Imagine shooting a film, then being able to Super impose anything you want in as fast as you can build it in mesh.

For example: I want to shoot a video of me eating a donut. But I don't have a donut. But I do have a mesh donut. Building the donut getting eaten would be a one off process which animators could easily change to any donut. Instead of worrying about the 3D space, let the AI compute this from video. Bind a hold from the donut to my hand and I can eat away.

This could really open up "anything is possible" in small amounts of time for film making.

There may be infinitely many ways to apply Tesla FSD to real world application.

Posted In:

jsturgis