The former MI6 chief Sir Richard Dearlove is no stranger to the intricacies of intelligence. But it was his comments about intelligence of the artificial kind, and about science, on the One Decision Podcast, which he co-hosts, that generated a flurry of interest last week. Reacting to grave warnings from some scientists over potential dangers of AI, Dearlove professed scepticism, reasoning that such dire predictions cannot be taken overly seriously given the failures of scientists on Covid. Such comments betray a common and insidious confusion over what science is and how it should be interpreted, and risk emboldening scientific denialists.
Part of the error stems from a mistaken conflation of “science” and “scientist”. Science is not an arcane collection of dogma but an active and systematic method of inquiry. Science pivots on making testable predictions, which are updated as new findings emerge, to reflect the totality of evidence. Scientific positions are always transient, subject to revision when stronger evidence emerges. All scientific knowledge is provisional, therefore scientific advice is prone to change and can evolve at dizzying speeds during periods of intense discovery.
