You can use all kinds of algorithms in machine learning. This is a “uses a” relationship rather than an equivalence relationship, in this case. If I’m building a piece of furniture, I am a carpenter. I could employ the use of a hammer to help me build the furniture. The hammer is not a carpenter.

I think you can imagine that the machine learning approach in that video may also rely on various data structures including graphs, trees, etc, and perhaps many other things which are also not machine learning.