Orateur
Boris Bolliet
Description
Generative AI is about to lead to a paradigm shift in the way we conduct research, with tremendous potential gains in productivity, robustness and discovery. We are developing multi-agent systems to automate and optimize scientific workflows in data analysis, with a focus on Cosmology. In this talk, we will present our prototype cmbagent, a pre-trained-LLM-based system capable of reproducing complex and state-of-the-art cosmological data analyses with minimal human input, built with ag2. Such systems offer a glimpse into the future of research, where the main task for human scientists will be the orchestration of self-driving laboratories.
