Yes, we were lucky. We went throw Ed Discussion to post

I already have postponed my exam and give extensions to my homeworks since all the material is on Canvas. I thought that single-point of failures were something in the past :)

What I tried once is to ask "Review this paper". It produces a (completely incorrect) review, but what I got out of it is that we needed to cover these incorrect relationships and differentiate in the related work.

I am curious about your take about this, Shriram. If authors feed their papers through the main LLMs, they will see what these reviews will look like and they can be defensive about what the LLMs will identify as perceived (and potentially wrong) weaknesses.

Looking good ...

They also need an institute of "ideas that have been tried and did not work". They will not know what to choose from :)