This wasn't always the case, but now almost any research topic in machine learning that you can imagine is now being done MUCH BETTER in industry due to a glut of compute and endless international talents. The only ones left in academia seems to be: niche research that delves very deeply into h
