Late‑Stage Neopositivism

The digital echo of neopositivism, where machine learning models and big data dashboards replace theory and interpretation. Late‑stage neopositivism treats correlation as causation, prediction as explanation, and training sets as reality. It is the ideology of the data scientist who has never read Karl Popper. It promises to solve social problems by optimising variables, while never asking what the variables mean or who set them.

Example: “The predictive policing algorithm was ‘evidence‑based,’ but its training data came from biased arrests—late‑stage neopositivism, garbage in, gospel out.”

Late‑Stage Neopositivism by Abzugal May 5, 2026