People increasingly interact in algorithmically controlled environments. Be it social networks, e-commerce or connected services our experiences are curbed by what is possible. News usually refers to it as experiments, echoing science methodology. How so?
Recent news of Alibaba’s experimenting with social behaviour add to the pile. We have seen Facebook, Amazon, Apple and governments – China’s Social Credit System the most obvious reference here – employ behaviour modification at large.
Regardless of the political and social impact of such exercise, I want to draw attention to the sort of ‘experiment’ we are talking about. These experiments do not render falsifiable theories. Their objects are not conceptually part of a hypothesis. The data and logical arguments are not objectively verifiable, and there is no intention of any peer revision.
Beyond replacing social conditioning, pavlovian stimuli, and psychological subjects for operands, signals, reinforcements and pother jargon, yet in this case science will have to cope with new validity criteria.