Just judging from the volume of papers, it seems to me that there are more academics writing papers on "AI safety" and "AI ethics" than there are academics publishing research papers on actual AI. It's become one of the hottest topics among legal academics, philosophers, ethicists, and a variety of connected disciplines, in addition to its niche among some computer scientists, and the amount of work to get to a paper in these fields is an order of magnitude less than actually publishing technical research.