I was surprised by how sensible people’s preferences were. When one was better than the other, they preferred the better one.
When the stakes were very high they preferred humans. I am inferring that on high-stakes issues, they are more hesitant to believe claims about the machines. I may be wrong that about that.
While there can be some concern about the representativeness of his sample (Mechanical Turk users), the idea of finding a population anywhere that is sensible on any issue is a cause for celebration.
Perhaps there is hope for humanity after all.
Or perhaps I am reading too much into the results.
Full text for “Worse Than Human?” is available for those that want to check whether I have cause for enthusiasm.