Comments by "Jim Mcneal" (@jimmcneal5292) on "AI Can’t Reason. Should It Drive Cars?" video.

With people problem is not that they have low reasoning capabilities, but with the process itself and the fact that people dislike when way world works don't fit their moral values. More precisely, when trying to understand phenomenon and build a model for it, people first pick some moral values on emotional basis(i.e. those that "feel right" to them). Then they try to construct a model, but place restrictions so the model/explanation wouldn't be "immoral". Since their moral views are picked on emotional basis, there's no reason for actual explanation to be "moral". But upon encountering this problem, people usually prefer to add new mechanisms to the model/add justifications. When at the certain point evidence pile up and it's just impossible to add justifications, they simply choose to ignore the evidence they were presented with. In general, such problem can be fixed in two ways(but both seems unlikely for people to start practicing) - either be ok with reality being "immoral"(i.e. best solutions being immoral) or consider practice the more moral the better it works.
1