The Nature folks have said they plan to focus more on interdisciplinary work, which makes sense from a strategy perspective since a publication in a Nature journal is more weighty in other fields, e.g. Biology.
Also worth noting is that DeepMind publishes in Nature for the prestige.
The Nature folks have said they plan to focus more on interdisciplinary work
Best of luck to them, I guess. I hope most/all AI and Machine Learning researchers continue to (or begin to) publish pre-prints to ArXiv and/or submit to JMLR, JAIR, and other open-access journals.