Comment by telesoft

Humans are wrong all the time. Do we really want "human-like" AI? And maybe this is where everything falls apart: will a machine as smart as a human be prone to the same errors as us?