At 50% the bot is about at “human” level, because the test subjects effectively cannot tell the difference in the Turing test. It’s a flip of a coin, despite the dynamic inspections. At 70%, does that mean they are “more human than humans”?

At 50% the bot is about at “human” level, because the test subjects effectively cannot tell the difference in the Turing test. It’s a flip of a coin, despite the dynamic inspections. At 70%, does that mean they are “more human than humans”?

That‘s a very good question.
And it makes my point. We don‘t know beforehands what behaving human means. We‘re figuring this out.