When does learning from data work (math starting from basic probability)

This is a very strange article that has a “write only” feel to it. I don’t know who the audience for it is, but as someone who actually wants to learn the mathematics of machine learning starting with probability, my plan is to work through Kevin Murphy’s “Probabilistic Machine Learning: An Introduction” https://probml.github.io/pml-book/book1.html which looks excellent.

What if the only real answer is "when you make it work"?

I don't mean to be rude, but this reads like 0 human oversight Ai slop