Assessing the risks and benefits of using AI-generated images in scientific research

In a recent Nature article, Carissa Wong discusses the potential opportunities and challenges associated with using AI-generated images and videos for scientific research. The author highlights the mixed responses to AI-generated images within the scientific community and discusses how these images are already being used to illustrate methods in scientific papers, promote papers in social media posts, and touch up presentation slides. Wong notes that generative AI tools can benefit researchers by reducing the time it takes to produce images at a higher quality, but advises that image-generators are not yet able to create complex scientific figures with accurate text annotations and may produce inaccurate or fraudulent images. The author concludes that responses to AI-generated images are still mixed within the scientific community.

Nature (Acct Req)