Martin C. Martin wrote:
...
> But say that to most AI researchers, and they'll stare at you
> uncomprehendingly.  They want a well defined problem, such as using all
> users purchases at Amazon to suggest other purchases for a single user.

   A while back, a DARPA program manager (an agent person, at that), 
sent out the notice to his program that the textbook on agents that he 
wrote before moving to DARPA was available on Amazon.  The beauty of 
this was the "people who purchased this" recommendations, which started 
with "Clean Underwear".  He reported this and I subsequently checked 
and, sure enough, Amazon recommended that purchasers of his book would 
also like to purchase clean underwear.  I suspect this was the default 
for something that had no purchasers, showing the sense of humour of the 
programmers.  However, I have seen many other nearly as absurd 
recommendations from that type of AI.  Clearly, the absurdity arises 
because they do not model the real world, just data mine blindly.  Those 
recommendation systems clearly do not pass the Turing test.

-- 
Ray Parks                   [EMAIL PROTECTED]
IDART Project Lead          Voice:505-844-4024
IORTA Department            Mobile:505-238-9359
http://www.sandia.gov/scada Fax:505-844-9641
http://www.sandia.gov/idart Pager:800-690-5288


============================================================
FRIAM Applied Complexity Group listserv
Meets Fridays 9a-11:30 at cafe at St. John's College
lectures, archives, unsubscribe, maps at http://www.friam.org

Reply via email to