I just started reading this out of curiosity and it really feels like it could be a strong supplement for reading ML theory. Has anyone else given it a read? What are your thoughts?