Yes I do believe Modularity as part of the first Seed AI. Seed AI should then allow its own redesign and reorganization.
Your mentioning of the structure/dynamics might foretell of a new or modified AI compiler-concept generator-organizer-System Analyzer for algorithms, structures and dynamics operating within those modules. Code generators will be a major part of the "KEY" to the algorithms, structures and dynamics operating inside the modules. ---- Dan ---------------------------------------------------- >From : Ben Goertzel <[EMAIL PROTECTED]> To : <agi@v2.listbox.com> Subject : Re: [agi] Hard Takeoff Modules? Date : Fri, 20 May 2005 18:59:30 -0400 > > I'm not quite sure I understand the question. > > You're assuming that a "seed AI" (an AI capable of self-modification toward > a trajectory of unboundedly increasing intelligence) will have a > multimodular design, and asking for a list of the modules in this design? > > Such a decomposition exists in the context of the Novamente AI design, but > the modular decomposition itself is not *that* interesting, the key thing is > the algorithms, structures and dynamics operating inside the modules ... no? > > -- Ben > > ----- Original Message ----- > From: <[EMAIL PROTECTED]> > To: <agi@v2.listbox.com> > Sent: Friday, May 20, 2005 10:01 AM > Subject: [agi] Hard Takeoff Modules? > > > > > > Does anyone have a list of Seed modules for a Hard-Takeoff > > or know of a web site that has a list? > > > > > > ---------------------------------------------------- > >>From : Lukasz Kaiser <[EMAIL PROTECTED]> > > To : agi@v2.listbox.com > > Subject : Re: [agi] Is "computation" a good concept for describing AI > > software? > > Date : Fri, 20 May 2005 15:43:40 +0200 > >> Hi Ben. > >> > >> > I've been talking with Pei Wang about whether "computation" is a good > >> > concept for modeling or describing AI systems. > >> > >> I'm not sure I understand you. When you are talking about > >> "computation" you seem to think about something roughly > >> similar to Turing machines. But you know very well that the > >> concept of "computation" is very broad with dozens of models > >> for parallel and / or probabilistic computation. And it seems > >> there are really strong arguments that these models are far > >> better for describing AI than the standard Turing machines, > >> that is hard to refute. But I can't see the link with special > >> relativity theory at all, sorry. > >> > >> - lk > >> > >> ------- > >> To unsubscribe, change your address, or temporarily deactivate your > > subscription, > >> please go to http://v2.listbox.com/member/[EMAIL PROTECTED] > > > > ------- > > To unsubscribe, change your address, or temporarily deactivate your > > subscription, > > please go to http://v2.listbox.com/member/?listname¢.listbox.com > > > > > > > ------- > To unsubscribe, change your address, or temporarily deactivate your subscription, > please go to http://v2.listbox.com/member/[EMAIL PROTECTED] ------- To unsubscribe, change your address, or temporarily deactivate your subscription, please go to http://v2.listbox.com/member/[EMAIL PROTECTED]
