[logancg]

How do you come to the conclusion that ML is a buzzword here? It's natural to publicize interesting new research that grounds a field -- as statisticial learning theory does for machine learning. Historically, Ben-David and co-authors have produced foundational work on the implications of SLT in ML.

[dooglius]

The problem is that the model was invented in this paper itself, and doesn't actually seem to relate to any actual ML algorithm. Admittedly, I didn't read the paper itself so this may be the overreach on the article's part.

[zwaps]

I don't think that can be true. If it was, it were like this:

Hey guys, here's an algorithm. Oh by the way, we can not prove that this algorithm does something useful, and their can you.

gets published in nature

Talk about bang for the buck