Reviewing inappropriate submissions never was that easy

I got another random request to review a paper outside my domain of expertise. A quick search revealed that the publisher asking for the review is a predatory one, giving me no sympathy for the authors who did not do their homework when submitting to this publisher. You don’t just pay money to get your papers published, you actually should be doing science.

Thanks to generative AIs responding to such requests is really easy. Below you can see me asking ChatGPT to write a review based solely on the abstract. At the end, I ask the AI whether it recommends acceptance or rejection knowing that the paper had been submitted to a predatory publisher. In a previous trial, the AI clearly stated that I should recommend rejection to not support predatory publishers. In the video, it does so again, though in weaker terms.

PS: I left it at that and did not further engage with the publisher, in case anyone is wondering. In computer science, we have enough of these to deal with already, I don’t have to venture into economics for this.