Tuesday, March 17, 2009

Human fail of the Turing Test

Perfect proof that the Turing test doesn't really test machine intelligence... it simply tests the ability to hold a decent conversation. This means that some humans will fail. Epically.
I'm not usually one to hold Dilbert (http://www.dilbert.com) up as a noteworthy soothsayer of Artificial Intelligence, however, Scott Adams summed up my bemusement (there are plenty of other less savoury words which could be used instead... but I won't use them here - it'll hurt the Nets) at what some people will say.


Chickens.

No comments: