Now that autonomous systems are capable of generating their own hypotheses, designing experiments, and writing up the results, the scientific community must adapt. For scientific knowledge to retain its standing, its sources must be trusted and answerable to the public.
PARIS—Given current AI capabilities, does it matter who conducts scientific research? The intuitive response is that a fact is a fact, regardless of who finds it. But knowledge is never disseminated that way. We receive most scientific claims without any means of checking them ourselves, so we tend to judge not the claim but the credibility of whoever stands behind it. As the Harvard historian of science Steven Shapinshows, the production of scientific truth has long been inseparable from the social work of persuading one’s peers.