An AI system at Chalmers now proposes, tests, and validates new biological hypotheses without human intervention. The team published results last month showing the system identified novel gene interactions in yeast, confirmed through wet-lab experiments at the university’s own facilities. It runs on standard compute, not supercomputers, and iterates faster than a human-led lab can schedule experiments. This is not a tool for researchers; it is a researcher that happens to be code. For Nordic builders, the leap is from automation to autonomy in discovery workflows. The implication is clear: labs that integrate such systems will outpace those that do not. This week, audit one internal process where hypotheses are still hand-crafted and map the data pipeline needed to let an AI scientist take the first pass.

researched · 3 sources

29 OctResearchreaches nearby