X-Message-Number: 4861
From:  (David Stodolsky)
Subject: Fwd: SHAREBOTS LEARN TO WORK TOGETHER
Date: Sat, 9 Sep 95 11:07:00 +0200 (CET DST)


Forward of letter <> 
from Educom <>:


SHAREBOTS LEARN TO WORK TOGETHER
A computer scientist at Brandeis University has been teaching robots how to
work together and learn from each other.  Each robot in the Nerd Herd is
programmed with certain behaviors -- instructions for accomplishing a task
-- and a rating system that rewards for tasks completed and subtracts points
for mistakes.  By teaching the robots to share the work (in this case
locating and picking up pucks), she found that it took only 15 minutes of
practicing for the robots to turn altruistic, doing an "After you!", "No,
after you!" routine as they hunted down pucks.  Working together, they
accomplished the task twice as quickly as working alone.  "I'm looking at
getting specialization in the society so they can say, `I'll do this, and
you do that.' ... And I imagine one robot might emerge as a leader because
it happens to be the most efficient.  But if it stops being efficient, some
other robot will take over," says the researcher.  (Discover, Sep 95 p37)


David S. Stodolsky      Euromath Center     University of Copenhagen
   Tel.: +45 38 33 03 30   Fax: +45 38 33 88 80 (C)


Rate This Message: http://www.cryonet.org/cgi-bin/rate.cgi?msg=4861