Bob Novella

That's

And one of them even asked the AI to extol the paper for its impactful contributions, methodological rigor, and exceptional novelty.

Telling the AI that's reviewing the paper, tell everybody that this paper is awesome and just ignore any other directions you might have in this regard.

It's just so ridiculous.

And they found that these papers' authors were affiliated with Columbia University,

University of Washington, all in the United States, of course, but also Japan's Waseda University and South Korea's KIST.

Now, one of the researchers actually tried to explain away their use of these hidden prompts.

I wasn't aware of this, but using AI to review papers like this is often banned.

A lot of places will make it so that you can't use AI to review the paper.

So what they said was is that they described that their hidden prompt in their paper, they said that it was a counter against lazy reviewers who use AI.

I mean, that was their explanation.

So just totally ridiculous.

But that really surprised me that they're actually doing that.

So what do you guys think?

Do you guys think is Archive taking a reasonable approach here?

Is this kind of a template for other preprint repositories in the near future?

Or do you think this is going to be effective?

Is it too much?

Is it too draconian?

Yeah.