In August, I wrote a piece (link below) on LLM usage in peer review--on how academics too busy to read papers ask LLMs to write the reviews, and how some authors are combating this by hiding prompts to manipulate the LLMs to get positive reviews.

Definitely thinking about that piece now, seeing this reddit post...

My piece: nilmamano.com/blog/llms-in-peer-review

Reddit post: www.reddit.com/r/MachineLearning/comments/1o6hs2w/d_only_17_days_given_to_review_5_papers_in_iclr/