Researchers criticize AI-generated figures in arXiv papers
Researchers on X have highlighted a growing aversion to AI-generated figures in new arXiv papers. Jiafei Duan observed that an opening AI-produced image often prompts readers to close the tab immediately. The discussion noted AI's role in accelerating research while urging authors to create their own figures. Robotics researcher Chris Paxton at Agility Robotics added that current tools remain limited for conveying ideas visually despite gains in coding and documentation tasks.
#755Chris Paxton@CHRIS_J_PAXTONFor all the progress in AI coding (and documentation and coding-adjacent stuff) they are still not great at communicating your ideas to someone else imo
Jiafei Duan@DJiafeiIs it just me, or every time I open a new arxiv paper and the first figure is AI-generated, I instinctively close the tab? AI is accelerating research, but please make your own figures.
Unless your ideas are mid as hell i guess, they do work for emails
Chris Paxton@chris_j_paxtonFor all the progress in AI coding (and documentation and coding-adjacent stuff) they are still not great at communicating your ideas to someone else imo