tag:blogger.com,1999:blog-9818962.post1758682507694474395..comments2024-03-20T13:10:11.477+05:30Comments on nanopolitan: Talent, Expertise, ExcellenceAbihttp://www.blogger.com/profile/06790560045313883673noreply@blogger.comBlogger1125tag:blogger.com,1999:blog-9818962.post-3623621383003869392009-11-30T01:15:15.895+05:302009-11-30T01:15:15.895+05:30You have a giant network running two learning algo...You have a giant network running two learning algorithms in parallel (mostly), a long-term one (evolution) and a real-time one (within-lifetime learning). Now, some of the parameters of the network comes preset in some organisms, in a way that helps them learn some things faster. Other organisms need to learn more to find those parameters. It is totally unclear to me why this is such a big issue.<br /><br />What is really interesting is the interaction between the two networks -- the interaction between genetics and plasticity -- and we don't know much about this. Though epigenetics looks like a promising window. <br /><br />Closely following is bad learning algorithms, such as allergies and stereotyping.<br /><br />I think the real insights in both these areas will come from modeling, not from the thrashing about of people like Gladwell.Anonymousnoreply@blogger.com