• We argue that the key bottleneck in science is the research paper as a medium for communicatingand storing knowledge. This inefficient format shapes the incentive structure in academia and isthe major cause of the reproducibility, serials, and peer review crises.• Reforming the publishing system is difficult because most academic performance metrics revolvearound publishing data as academic papers. Changes must be evolutionary and seamless.• We propose that communicating data as nanopublications connected within an argumentationgraph can be a more efficient way to communicate knowledge to both humans and AI algorithms.This format can create an entirely new and efficient type of interaction with knowledge, alsohelping to resolve the pressing crises.• A successful reform of scientific communication requires migration of substantial amounts ofdata from research papers into a new format. Current AI technologies enable converting largevolumes of documents, making the transition to a new publishing practice feasible.
Lidsky et al. (Wed,) studied this question.