10-17-2010, 08:31 AM
If the first law is Do not harm humans (or by in action cause any humans to come to harm), a robot has to know when it is harming humans:
<!-- m --><a class="postlink" href="http://www.newscientist.com/article/mg20827826.700-robot-arm-punches-human-to-obey-asimovs-rules.html">http://www.newscientist.com/article/mg2 ... rules.html</a><!-- m -->
--tg
<!-- m --><a class="postlink" href="http://www.newscientist.com/article/mg20827826.700-robot-arm-punches-human-to-obey-asimovs-rules.html">http://www.newscientist.com/article/mg2 ... rules.html</a><!-- m -->
--tg