I was debating whether to call it a modified Turing test in the title, but had already editorialised it quite a lot. What was interesting though is that the bo

I was debating whether to call it a modified Turing test in the title, but had already editorialised it quite a lot. What was interesting though is that the bots and humans did not converge on optimization strategies, and part of appearing human was programming irrational behaviour (such as grudges) into the bots.