I was thinking the exact same thing - sadly ML folks stopped caring about existing acronyms a while ago - I've seen multiple papers in ML with the same acronym. People want something catchy for many reasons (to make it more popular and get more citations for the most part) and there are only so many catchy acronyms one can come up with.