Tech Review writes about fascinating work creating a test to determine if a bot is human. The original Turing Test came from a proposal by computing pioneer Alan Turing to demonstrate machine intelligence by fooling humans into thinking that a computer was another human during a question and answer process.
This new approach is a contest that ' ... challenges programmers to create a software "bot" to control a game character that can pass for human, as judged by a panel of experts. The goal is not only to improve AI in entertainment, but also to fuel advances in non-gaming applications of AI. The BotPrize challenge is a variant of the Turing test ... ' .
What is most interesting thing about the contest is that it deals with physical activities of the bots as they would be seen in gaming contexts. Natural language is not considered. Since language is such a big part of our detecting intelligence, this would appear to be a very different kind of test. Still, physical intelligence is also worth measuring.
Here is a summary of the 2009 contest which provides summary results. Later details will be posted. This was part of the just completed 2009 IEEE Symposium on Computational Intelligence and Games. Here are some video clips of the contest tests.
As in the Loebner Prize for the standard Turing test, no machine has yet been perfect in fooling all of a panel of experts into thinking it was intelligent.
No comments:
Post a Comment