Humans (at least most) have a sense of fallibility. Thus when unsure we do not simply go with the best assumption. "AI"s seems to just go with whatever is the

Humans (at least most) have a sense of fallibility. Thus when unsure we do not simply go with the best assumption. "AI"s seems to just go with whatever is the "most" likely, even if it is just some 0.0001% above a sea of alternatives.