Academics accuse AI startups of co-opting peer review for publicity

A controversy has emerged at the ICLR conference over AI-generated studies submitted by labs Sakana, Intology, and Autoscience. While Sakana informed ICLR leaders and obtained consent from peer reviewers before submitting its AI-generated papers, Intology and Autoscience did not, leading to criticism from the academic community. Critics argue that using peer review as a benchmark for AI technology without consent undermines the scientific process and disrespects the time and effort of human reviewers. The issue highlights the growing problem of synthetic text in academic submissions, with a significant percentage of papers at AI conferences containing AI-generated content.

The situation has sparked discussions about the need for regulated evaluations of AI-generated studies, with suggestions that researchers should be compensated for their review efforts. Sakana admitted to errors in its AI-generated papers and withdrew its submission to maintain transparency and respect for conference norms. The incident underscores the challenges academia faces with the increasing use of AI in research and the ethical considerations surrounding the peer review process.

Key takeaways

AI labs Sakana, Intology, and Autoscience submitted AI-generated studies to ICLR workshops, with only Sakana informing ICLR leaders beforehand.
Critics argue that using peer review as a benchmark for AI-generated studies disrespects the time and effort of human reviewers.
Sakana admitted to errors in its AI-generated papers and withdrew its submission, while Intology received positive reviews but faced criticism for not disclosing AI involvement.
There are calls for a regulated agency to evaluate AI-generated studies, with researchers being compensated for their time.

Academics accuse AI startups of co-opting peer review for publicity | TechCrunch

Key takeaways

Discussion (0)