Hacker News Reader

>"Ten years ago, if I had this conversation, I would have assumed the entity on the other end was a human. You can no longer take it for granted that an AI does not know the answer to “common sense” questions."

[...]

>"Now we’re getting into surreal territory. GPT-3 knows how to have a normal conversation. It doesn’t quite know how to say “Wait a moment… your question is nonsense.” It also doesn’t know how to say “I don’t know.”

"GPT-3, why do two wrongs not make a right -- but three lefts, do?" <g>

Good read, but I wish the author would have conducted a better Turing test by having someone that does not know beforehand interface with the AI. Add a control, find more people for both acting as control and participating in discrimination, and it would have been fascinating to see the results of such a test.

I wonder if it's just coincidence that all the "heavier" answers are whatever was listed second.

I don't think it's just coincidence. A lot of "A or B" questions tend to have B as the correct answer with the first option just being there to confuse the reader. I find it likely that a statistical algorithm would pick up on that.

Giving GPT-3 a Turing Test