Path: utzoo!utgpu!jarvis.csri.toronto.edu!rutgers!aramis.rutgers.edu!athos.rutgers.edu!nanotech From: alan@oz.nm.paradyne.com (Alan Lovejoy) Newsgroups: sci.nanotech Subject: Re: Swamping AI-based design with computational complexity Message-ID: Date: 1 Jul 89 02:20:28 GMT Sender: nanotech@athos.rutgers.edu Organization: AT&T Paradyne, Largo, Florida Lines: 27 Approved: nanotech@aramis.rutgers.edu In reply to article dmo@turkey.philips.com (Dan Offutt) JoSH writes: <[The problem of finding *the* *optimal* active shield is surely as < intractable as you describe. However, these same arguments apply to < writing optimizing compilers, or playing chess. This doesn't mean that < the problem of writing a useful optimizer or playing an acceptable < game is intractable. It just means that designing active shields < is like the rest of life, with no absolute guarantees of success. < --JoSH] Dan's arguments against active shield design apply just as well to the design of gray goo. Often, computationally-intractable problems can be solved by finding a better algorithm--or by adopting an approach which does not depend on finding the optimal solution--or any solution--to the intractable problem. For instance, using traditional methods on a digital computer, it is VERY difficult to build a system that can recognize human faces. But now we know that there are much better ways. What new mathematical techniques might be discovered by an ultra AI? Alan Lovejoy; alan@pdn; 813-530-2211; AT&T Paradyne: 8550 Ulmerton, Largo, FL. Disclaimer: I do not speak for AT&T Paradyne. They do not speak for me. ______________________________Down with Li Peng!________________________________ Motto: If nanomachines will be able to reconstruct you, YOU AREN'T DEAD YET.