Can we pinpoint a specific decade this happened or is it a slow process? Because if we count CS as a science, I’m seeing it become just like the worst of the sciences in this regard. AI has really gotten me pessimistic. There’s so much borrowing from other subjects since there are now models for every subject and every grade level, which ultimately will become products for large corporations. Like maybe we should hire at least one full time linguist, mathematician, chemist, logician, etc to sit high in the chain behind each of the respective models. I’m seeing a lot of faux rigor because it’s moving lightning speed.
I've only had an inside view of academia since the mid aughts, but it was already a well established issue then with departments trying to institute p-hacking controls and citation rings being a known thing. I think a big part of why it's exploding so hard is the scarcity of academic jobs combined with the big push to increase STEM PhD production leaving people to a cutthroat competition, where you have to either cheat or be absolutely brilliant to have a real shot at a tenured position.
