The big issue with exploiting massive data sets for social engineering was that they reflected too much inconvenient reality, ill-suited for useful narratives so long as the firehose of unfiltered stats prevailed. The solution, many specialists believed, was algorithmic innovation—honing new means of spiking problematic data or at least boosting the better stuff to feed more progressive machine learning.

With ever-sharper artificial intelligence thus trained for equity-minded global technocracy, disregarding bad data, multidisciplinary computer scientists could turn to an even more ambitious vision: systematically solving inequality via quantified intersectionality putting thumbs on every scale at once.

In accounting for race, gender, income, et al programmers were gaining the power to construct new hierarchies at whim. It was becoming startlingly easy to install novel schemes of advantage and disadvantage, rewards and sanctions, like assigning magic points in games of Final Fantasy.

Nobody ever used the term “social credit system,” but the implications were clear …

[And getting clearer all the time. — Ed]