Turing Test – 2014

By on August 12th, 2014 in Human Impacts

Chatbot “Eugene Goostman“, created by a team of Russian & Ukrainian software developers: Vladimir Veselov, Eugene Demchenko, Sergey Ulasen has “passed” the Turing Test (as of June 6, 2014).  I’d like to report my own interactions with Eugene, but the links a current instantiation seem to be ‘temporarily’ disrupted — I suspect for purposes of monetizing the recent notoriety.

Ray Kurzweil, in his 1999 book “The Age of Spiritual Machines” predicts that  by 2019 there would be “widespread reports of computers passing the Turing Test although these tests do not meed the criteria established by knowledgeable observers.”  It appears that Ray is right so far with claims of earlier successes, and deprecation of this event as being sufficient.

I won’t duplicate notes of prior posts on AI, but will point out that “practical” applications of Chatbots and other AI type software exist and will have ‘social impact’.  One impact will be the expansion of online interaction that can provide useful responses to consumers, students, etc.  Fooling the public is not needed (and may be unwise) … at times having a clearly ‘computer generated voice’ (audio or otherwise) helps set the expectations of the humans interacting with the system.  However, we can expect increasingly sophisticated capabilities along these lines.

What uses/services would you suggest as priority applications for a fairly robust ‘chat-bot’?