Could you motivate high school math by using it to explain how AI works?

More interesting? I doubt it.

Linear algebra, probability, and symbolic logic require significant prior work with math that high school students usually don't have exposure to. Getting them through arithmetic and algebra apparently presents enough problems.

You’re going to need at least calculus for derivatives and probably also some linear algebra which is rare to get in high school. Most high school students never get to calculus and of the ones that do there or in college, a lot of them struggle with derivatives, never mind integrals.

Derivatives are already motivated by geometry, physics, and optimization. Aren’t neural nets just another form of optimization?

Adding any kind of practical application to high school math would make it more interesting. It doesn’t have to be AI, it could be car design. Math is usually taught abstractly with no connection to the real world. It’s no surprise students aren’t interested.