Biomedical Writing Consultant ◆ Scientific & Medical Writing Instructor ◆ Manuscript Editor ◆ Grant Editor ◆ Award-Winning Educator ◆ Writing Coach ◆ Speaker ◆ Scientist
"Although the current ability of ChatGPT to generate experimental images is limited, our simple tests have demonstrated the significant risks of misuse in generating images. . .ChatGPT theoretically has the potential to generate entire academic papers from scratch, including text, raw data, and experiment result images. While images generated by ChatGPT currently are not as realistic as those generated by GANs [Generative Adversarial Networks], the low barrier to use and rapid technical improvements mean the generated images will likely be more realistic in future." What do you think about using ChatGPT to create images for academic papers? #medicalwriting #science #academia
Due to the inaccessibility and high cost of BioRender, it is beneficial to have a budget-friendly alternative that allows for the inclusion of more schematics and visuals in the manuscript. Visuals enhance readability, and the more images included, the better.
I haven't tried ChatGBT for images of any kind Crystal Herron, PhD, ELS. I have been using Adobe Firefly for images for my blog, slides for presentations, etc. It can create fantastic images for a fairly low monthly fee. But the images I'm working on are for marketing, not scientific images. I don't know how Firefly works for that.
AI is powerful & helps to reproduce data and is made by humans with available information whereas humans are intelligent by nature, so will be ahead of AI.
Freelance science writer helping life science companies with content marketing to attract and engage their audiences. Microbiologist.
1moThose images look too clean to be real and it scares me how AI could be use for fake data, writing publications, etc. But I wonder about it's use to generate more schematic figures. Like I often draw out schematics type figures by hand (and they are ugly!), and try to recreate them in PPT, Canva, etc. for presentations, etc. 😬 Maybe that capability exists already, but being able to input a hand drawn schematic and have it spit out something more polished could be an application that I find not as worrying.