TLDR - both papers got flagged and rejected by automated integrity checks.
VCFundedGenYer•42m ago
Please, even if it is an experiment, don't do this. We don't need more AI slop in the world.
pizza234•27m ago
This experiment was agreed with the managing editor of the multidisciplinary and open-access journal IEEE Access, and IEEE itself - it was red teaming, essentially.
jszymborski•41m ago
Sidenote: way to flex your "writing studio".
kirth_gersen•33m ago
Save yourself a slog. The automated checks found that the citations didn't exist and flagged the "papers". A better title would have been "Automated Systems seem to hold back the AI Slop papers. For now." Sweet writing studio, though.
dist-epoch•28m ago
There is still no paper or notable result in gender studies published by an AI.
scotty79•21m ago
In other news people prompted a model to pick a problem to solve, solve it and write a paper about it. And it did and the paper contains novel contributions.
paxy•54m ago