An AI system at Chalmers University of Technology now generates and validates its own scientific hypotheses in biology, without human intervention. The team published the method on 2026-09-26, showing the system identified a previously unknown enzyme variant that improves a key industrial process by 12 percent. It then designed, ran, and interpreted the experiments to confirm the result, end to end. This is not automation of known workflows; it is a closed loop where the network proposes, tests, and refines its own ideas. For Nordic builders, this signals a shift from AI as a tool to AI as a co-author of new knowledge, with immediate implications for life sciences, materials, and energy research. This week, audit one internal research pipeline for steps that could be handed to an autonomous loop, then pilot it on a low-risk, high-iterative task.

researched · 3 sources

26 OctResearchreaches nearby