Reactions from ranked influencers
3 posts
the principle of deciding authorship regardless of whether a participant was an AI or a human is a good one IMHO. The only funny thing about this is that the AI system is not listed as a co-author. I guess one day, we will also want to change this. Interesting times.
@pli_cachete @jdlichtman I would just be honest about the distribution in this case. I think if you had a human collaborator and had done this portion of the work you would obviously be a coauthor!
@CsabaSzepesvari In old ages, part of the job of a mathematicians was to calculate (pi, e, log, trig, ...) and they publish them as an author. Digital computers made that task irrelevant. People just mentioned "we used ENIAC/PDP9 (or LAPACK or numpy)" without assigning authorship to the computer.
@CsabaSzepesvari I think the same should/will happen to AI assisted/generated proofs as well. They should be credited as tools, not as co-authors.