There is often interest in the results of machine learning algorithms, specifically because they came from machine learning algorithms. It seems similar to the ‘from the mouths of babes’ trope, where comments made by children are sometimes regarded as containing some special wisdom, due to the honest and innocent nature of their source. Similarly, people seem to think that the impassionate learning of a machine might extract some special insight from a data set.
(Of course, anyone with such opinions has obviously never met either a machine learning algorithm or a child.)
Has this effect been discussed or studied anywhere? Does it have a name?