We know people can lie. We also know that some people are more prone to lieing. Using these priors we can navigate daily uncertainty.

This sort of article/post helps us (as a society, not just HN) adjust those priors re AI. Because "computer says no" is a real thing.