hi, > > Ahhh, all this brings back memories of the good old days > at Webmind... :) > > It's just a pitty that Jeff isn't here :( >
You know, I've felt that SO often and so intensely lately, as Izabela and I have been working out all the kinks in PTL [a note to outsiders: Jeff pressing was a colleague of Shane's, pei's and mine at webmind Inc. in 1997-2000. jeff was a Webmind Inc. cofounder, and an amazingly brilliant scientist and human being. he died of a sudden, freak meningitis infection last year ... a terrible, terrible loss ;-< ] PTL is a probabilistic inference system that jeff and I worked out at Webmind, with help from pei and shane and others. The PTL that Izabela and I have now, however, has only one equation in common with the one jeff and I originally worked out. But we have finally made it work -- selforganizing probabilistic inference in an integrative AI context ;-) It is very sad that we can't share all our discoveries with jeff. Unless Tipler is right and he'll be back eventually ;-p > > But the structures and dynamics needed to make intelligence happen under > > reasonable space and time resource constraints -- THESE, I believe, > > necessarily involve primary theoretical constructs VERY DIFFERENT FROM > > computation theory, which is a theory of generic computational processes > > This I consider an open question. My belief is that many of the things > we now consider to be practical and engineering motivated do in fact > arise from foundational theory interacting with various time and space > complexity constraints. Which is, I guess, the Holy Grail of complexity > theory in a nut shell. > > Even if this isn't the case (or can't be proven) I also believe that > foundational theory on the nature of intelligence and results that > follow from that will offer significant insights into the practical > construction of artificial intelligence. Hopefully work by myself > and others in my group at IDSIA should help answer some of these > important questions over the coming years! This is where our intuitions continue to differ, I suppose. I think your statements *might* be correct if you interpret "foundational theory" to mean a very DIFFERENT foundational theory of computing than any that we now possess. But I don't think that current computing theory, even alg. inf. theory, is anywhere close to adequate to make pragmatic AGI stuff emerge as "foundational theory interacting with various time and space complexity constraints." Incremental improvements to current computing theory, as Schmidhuber and Hutter at IDSIA (and others) are doing will never get you there. Maybe you'll be the one to throw all that out in favor of a new computing theory that will be adequate in that regard. In my view, you should start off by taking pattern, not program or machine, as the basic elementary concept ;-) ben ------- To unsubscribe, change your address, or temporarily deactivate your subscription, please go to http://v2.listbox.com/member/?[EMAIL PROTECTED]
