Conclusion
Scientific and technical monitoring - Leveraging generative artificial intelligence

Generative AI technologies are profoundly transforming the practice of scientific and technical monitoring. In just a few years, since the public announcement of ChatGPT, tasks that used to take days to complete are now automated with high-quality results.

But this technological power must not obscure the gray areas. Hallucinations are not a minor issue; they pose a fundamental problem: how can we trust a system that can assert factually false statements with as much confidence as if they were true?

Furthermore, issues such as the overrepresentation of English, certain geographic regions, or the most well-known publications reinforce the inequalities already present in global scientific research. In addition, the lack of reproducibility of results also calls into question standards of scientific rigor, and the lack of transparency in the...