ChatGPT-based representation of the Tagus River through the river putative microbial profile
This content is an early or alternative research output and has not been peer-reviewed by Cambridge University Press at the time of posting.
Comments
"This is a fascinating interdisciplinary study that effectively bridges microbiology, AI, and design. The idea of using ChatGPT to interpret and communicate microbial ecological services to a non-expert audience is innovative and timely. I appreciate the clear methodology, especially the use of nanopore sequencing and the thoughtful approach to handling basecalling errors. I have two questions for the authors: Given that ChatGPT’s training data may not include the most recent or niche ecological research, how did you validate or cross-check the ecological services it generated (e.g., against known databases like KEGG or MetaCyc)? The study mentions that the 'Organic matter decomposition' service suggests urban residue presence. Could this inference be strengthened by correlating microbial data with physicochemical parameters (e.g., BOD, nutrient levels) from the same sampling site? Also, a minor suggestion: it would be helpful to see a brief discussion on the ethical considerations of using AI to 'speak for' non-human entities, especially in participatory design contexts. Overall, this work opens exciting avenues for public engagement with microbial ecosystems and biodesign. Looking forward to seeing how this framework develops with expert validation and larger datasets."
