There were 60,000 (sixty thousand!) paper submissions to #ICLR2027 (15th International Conference on Learning Representations). This is crazy. Some are arguing for using AI for reviewing, which is completely insane. Maybe, instead, we should organize conferences by only inviting speakers. This will furthermore provide opportunities for better representation of women, researchers from low-income countries, etc. Those 60,000 papers can all go online to feed the slop machines into model collapse.
3 Boosts
0 Likes
@djoerd It’d be wonderful to have the means to treat these slopfrauds the same way as dopers in sports. Get caught once, get a 2 year ban from publishing and attending, get caugth twice, bye bye scientific career.
@djoerd That’s insane!
#ICLR2027 solution: "No author may appear as a co-author on more than 20(!) papers" and "if you have never had a paper accepted for publication in a major machine learning conference or journal, you may submit at most one paper" (which literally is gate-keeping the conference from new researchers) 🤦
@djoerd Oof… That seems backwards.
@djoerd
💔🇵🇸 I’m Hadeel, a mother in Gaza. My husband has an amputated foot and cannot work, while my children, Hamza and Jouri, face hunger and fear in a fragile tent. I’m struggling alone to provide even the most basic necessities.
Please don’t leave us to suffer alone. My children need food, and our family needs your help to survive. 💔
Donation link on my page