Lots of papers, specially in the "AutoML" community are calling "active learning" as "human-in-the-loop", without a single reference to active learning papers.
[I will not publicize them here, but I can mention some of them if asked.]
I think umbrella terms like "active learning", "deep learning", "automl", etc. don't help* much for real science, but renaming them is even worse.
Additionally, under such terms, solutions from the past century are rebranded as new.
Regards