LLM AI: Scientific Discovery using Literature & robot scientists

The first approach is “literature-based discovery” (LBD), which utilizes language analysis, similar to ChatGPT, to analyze existing scientific literature. Its goal is to uncover new hypotheses, connections, or ideas that may have been overlooked by humans. LBD has shown potential in suggesting new experiments and research collaborations, promoting interdisciplinary work, and sparking innovation across different fields. Additionally, LBD systems can pinpoint areas where research is lacking, predict future discoveries, and identify the researchers likely to make them.

The second area involves “robot scientists,” also referred to as “self-driving labs.” These are robotic systems that leverage artificial intelligence to generate new hypotheses by analyzing existing data and literature. They then validate these hypotheses by conducting numerous experiments, spanning fields like systems biology and materials science. Robot scientists, unlike their human counterparts, are less influenced by prior results, exhibit reduced bias, and are highly replicable. They have the potential to expand experimental research, uncover unforeseen theories, and investigate avenues that human researchers may have overlooked.