Michael> On 10 Nov 2006 at 13:48, Eric Baum wrote:
>> I am arguing that if you find a code that solves a range of
>> naturally presented problems, that is so constrained (for example
>> by being extremely concise) that the only way it could have existed
>> is if it is based on an ingenious abstraction hierarchy with
>> modules exploiting real structure in the world and reused in
>> different contexts to solve different problems, then it will
>> continue to solve most new problems,

Michael> Ok, that seems pretty obvious, at least for anyone who
Michael> accepts the basic premise that the real world is highly
Michael> ordered. I don't think you're saying anything controversial
Michael> there, though people who don't like the idea of AIs being
Michael> based on serial-looking code will probably nitpick 'range of
Michael> naturally presented problems' in the hope of demonstrating
Michael> that only fuzzy/connectionist/massively parallel code can
Michael> solve nontrivial challenges (I'm not one of those people).

Michael> Generalising to this level isn't really saying anything
Michael> useful though, as you've pushed all of the challenge into
Michael> finding that code/module set/abstraction hierarchy in the
Michael> first place. 

That's a reasonable worry, but I think you get a fair amount out of
it. It tells you that much of the guts of cognition is in the genome, 
you shouldn't expect a tabula rasa algorithm or any other weak method
to solve the problem. It suggests
that neither hand-coding nor simple learning will produce
intelligence-- you will have to find a way to guide computers to
produce the code. It says quite a bit about what the code should look
like. It gives a different picture on what language is, and on why 
humans are cognitively so different from chimps. etc etc. 

-----
This list is sponsored by AGIRI: http://www.agiri.org/email
To unsubscribe or change your options, please go to:
http://v2.listbox.com/member/?list_id=303

Reply via email to