“Many machine learning models generate their results by operating on high dimensional correlations that are beyond the interpretive capabilities of human scale reasoning. In these cases, the rationale of algorithmically produced outcomes that directly affect decision subjects remains opaque to those subjects. While in some use cases, this lack of explainability may be acceptable, in some applications, where the processed data could harbour traces of discrimination, bias, inequity, or unfairness, the opaqueness of the model may be deeply problematic.” (Leslie, 2019, p. 4-5)