Terren -- The AGI-design ideas you mention are pretty much the place I started in the 90's, and described in Chaotic Logic ...
Gradually however I realized that to achieve advanced AI with currently-realistic computational resources, it's necessary to impose an overall structured cognitive architecture and then let this kind of autopoietic creativity occur **within the context of the architecture**, rather than trying to get the architecture to emerge. Note that in biology, evolution learned the architecture, but cognitive autopoiesis/creativity occurs within the (highly constrained/biased) framework of the evolved architecture.. -- Ben G p.s. I wish people would reply-as-text not as HTML ;-) On Fri, Oct 10, 2008 at 12:55 PM, Mike Tintner <[EMAIL PROTECTED]>wrote: > Terren, > > Thanks for reply. I think I have some idea, no doubt confused, about how > you want to evolve a system. But the big deal re autopoiesis for me - > correct me - is the capacity of a living system to *maintain its identity* > despite considerable disturbances. That can be both in the > embryonic/developmental stages and also later in life. A *simple* example of > the latter is an experiment where they screwed around with the nerves to a > monkey's hands, and neverthless its brain maps rewired themselves, so to > speak, to restore normal functioning within months. Neuroplasticity > generally is an example - the brain's capacity, when parts are damaged, to > get new parts to take on their functions. > > How a system can be evolved - computationally, say, as you propose - is, > in my understanding, no longer quite such a problematic thing to understand > or implement. But how a living system manages to adhere to a flexible plan > of its identity despite disturbances, is, IMO, a much more problematic thing > to understand and implement. And that, for me - again correct me - is the > essence of autopoiesis, (which BTW seems to me not the best explained of > ideas - by Varela & co). > > > Mike, > > Autopoieisis is a basic building block of my philosophy of life and of > cognition as well. I see life as: doing work to maintain an internal > self-organization. It requires a boundary in which the entropy inside the > boundary is kept lower than the entropy outside. Cognition is autopoieitic > as well, although this is harder to see. > > I have already shared my ideas on how to build a virtual intelligence that > satisfies this definition. But in summary, you'd design a framework in which > large numbers of interacting parts would evolve into an environment with > emergent, persistent entities. Through a guided process you would make the > environment more and more challenging, forcing the entities to solve harder > and harder problems to stay alive, corresponding with ever increasing > intelligence. At some distant point we may perhaps arrive at something with > human-level intelligence or beyond. > > Terren > > --- On *Fri, 10/10/08, Mike Tintner <[EMAIL PROTECTED]>* wrote: > > From: Mike Tintner <[EMAIL PROTECTED]> > Subject: Re: [agi] open or closed source for AGI project? > To: [email protected] > Date: Friday, October 10, 2008, 11:30 AM > > Terren:autopoieisis. I wonder what your thoughts are about it? > > Does anyone have any idea how to translate that biological principle into > building a machine, or software? Do you or anyone else have any idea what it > might entail? The only thing I can think of that comes anywhere close is the > Carnegie Mellon starfish robot with its sense of self. > ------------------------------ > *agi* | Archives <https://www.listbox.com/member/archive/303/=now> > <https://www.listbox.com/member/archive/rss/303/> | > Modify<https://www.listbox.com/member/?&;>Your Subscription > <http://www.listbox.com> > > > ------------------------------ > *agi* | Archives <https://www.listbox.com/member/archive/303/=now> > <https://www.listbox.com/member/archive/rss/303/> | > Modify<https://www.listbox.com/member/?&;>Your Subscription > <http://www.listbox.com> > > ------------------------------ > *agi* | Archives <https://www.listbox.com/member/archive/303/=now> > <https://www.listbox.com/member/archive/rss/303/> | > Modify<https://www.listbox.com/member/?&;>Your Subscription > <http://www.listbox.com> > -- Ben Goertzel, PhD CEO, Novamente LLC and Biomind LLC Director of Research, SIAI [EMAIL PROTECTED] "Nothing will ever be attempted if all possible objections must be first overcome " - Dr Samuel Johnson ------------------------------------------- agi Archives: https://www.listbox.com/member/archive/303/=now RSS Feed: https://www.listbox.com/member/archive/rss/303/ Modify Your Subscription: https://www.listbox.com/member/?member_id=8660244&id_secret=114414975-3c8e69 Powered by Listbox: http://www.listbox.com
