https://www.nature.com/articles/d41586-024-04021-w
What could go wrong really?
We should really start pushing #replication studies, not novelty for the sake of novelty. Also, I thought we were past #novelty = #impact ...
@nicolaromano I agree. Utterly terrible idea. Never mind the fact that there is no commonly accepted definition of “novelty” within disciplines.
@daniel @nicolaromano I also agree. And there is also no commonly accepted definition of scientific or societal “impact” (e.g. result, output, outcome, effect) in research.
@jasemrau @daniel Agree! Also "The novelty score is calculated using an algorithm that compares the combinations of keywords and cited journals in a scientific manuscript with those in previous publications and projects the types of paper that will be published in the future". This is really calling for people to game the system by using buzzwords and citing high impact-factor (not necessarily high impact!) papers. Just need to get a list of high-novelty keywords... :/
@nicolaromano wow, didn't expect that from science. Pushing for novelty can be very bad and dangerous for innovation.
Take my field: drug design. Lately there has been a large application of different generative algorithms to the process. Very innovative articles come out every few days presenting a new algorithm to work with drug design.
None of these algorithms are actually ever used to design drugs. Sure, a bit of improvements over those methodologies is good before you apply them, but it's now been 7 years, I think it is time we test whether all these methods actually do work or not.
@nicolaromano
That's scary and makes me avoid publishing my paper somewhere use an algorithm to help them decide what to publish.
Search engines are already using some technique to score good content, and now if you want a good article to read, sometimes you'd better not use a search engine as in its 2024 sense. Because when you use an algorithm to sort out good content, a lot of people will work only to get selected by the algorithm, even if this leads to their content becoming far from good