AI scientist autonomously generates and validates new biological discoveries

Researchers at Chalmers University of Technology in Sweden have created an AI scientist that autonomously generates scientific hypotheses, designs experiments and interprets results. This groundbreaking development marks a significant step towards self-driving laboratories, where AI can conduct research on biological systems with minimal human intervention.

By harnessing advances in large language models, automated reasoning, and laboratory automation, the AI system was able to investigate brewer's yeast, Saccharomyces cerevisiae, by integrating its scientific knowledge, including the yeast's genome, metabolism, and prior studies.

The AI scientist demonstrates the potential to revolutionize research by rapidly exploring complex scientific questions and optimizing the use of laboratory resources. Ievgeniia Tiukova, a postdoctoral researcher and co-author of the study, emphasizes that while the AI can generate new scientific knowledge, it primarily serves as an augment to human scientists, taking over routine cycles of hypothesis generation and experimental testing.

Ross King, the study's senior author, explains that autonomous laboratories will collaborate with human scientists to accelerate discoveries in biology, medicine, and biotechnology, ultimately becoming increasingly capable partners in addressing challenging questions in these fields. However, humans remain essential for defining research priorities, interpreting broader scientific significance, and ensuring ethical oversight.