Industrial scientific paper mills
Statement (NASA form)
Given that models already draft passable papers, figures, and peer reviews, there is a possibility of journals and grants flooding with synthetic scholarship resulting in a literature no one can trust enough to build on.
Science is a reputation system. Flood it and it stops working.
Composite 15 = 4×3 + 3
Applicable mitigations
Controls · Hard fences on diagnosis, dosing, and triage language · Grounded, dated medical sources with uncertainty on the surface · Regulate high-stakes health modes as medical devices · Apprenticeship quotas in firms that deploy the models · Automatic stabilisers for exposed graduating cohorts · Transition income tied to measured displacement, not rhetoric