In response to a suggestion that PM made, I originally replied that differentiation would be easy. I guess that is not right. The reason is because the method of scalability that I will rely on will be correlation. I would be using the combinations of different types of *insights* or *ideas* anyway so these would act as correlations between different types of data objects that were somehow related. Correlation is a normal method of scalability in AI but I will not be limited to using statistical correlation in my program (although I will use it if I see a good reason to use it.) The problem with integration is the problem of integrating different kinds of relevant insight or information so the integration problem is also a problem that I expect to find with correlation. The solution to integration and to differentiation is to find out more about the objects being integrated or differentiated. I feel that this method is also necessary for judgement and awareness (in the basic sense). So all these ideas are closely related in my plan. However, what I wanted to say is that both differentiation and integration are going to be difficult.
Jim Bromer On Fri, Sep 14, 2012 at 9:42 PM, Piaget Modeler <[email protected]>wrote: > Building on what Stan said, you can also substitute "case" for reason as > well. Rules or cases, either will do. > > Jean Piaget talks about Differentiation and Integration as the hallmarks > of intelligence. > > Douglas Hofstadter has done a lot of work with analogy (which is one form > of integration). > Simple Crossover (from classifier systems) is another form of integration. > > PAM-P2 takes both approaches, having an analogizer component to perform > concept substitution according to specified "rules", > as well as concept invention by means of crossover during regulation (the > regulator component governs this). > > There are many ways to achieve integration though. The real trick is > differentiation. > > Cheers! > > ~PM. > > > ------------------------------------------------------------------------------------------------------------------------------------------------ > > > > Date: Fri, 14 Sep 2012 18:42:42 -0600 > > From: [email protected] > > To: [email protected] > > Subject: Re: [agi] Simplistic Test of Reason-Based Reasoning > > > > > Greetings Jim. > > my reaction to a couple statements contained below - > > > > > > On 09/14/2012 08:06 AM, Jim Bromer wrote: > > > > > > > > > I was wondering if a simple system of reason based reasoning could be > > > used to start an expanding system of knowledge acquisition. I am not > > > talking about a human-level AGI program. I am talking about a very > > > simple, very artificial system to test the viability and the > flexibility > > > of the reason-based reasoning strategy for general learning. > > > > stan: I like the beginning line, especially the word "simple." A > > bootstrap kind of program is exactly what is needed. > > > > To stick with the simple description, I would avoid the word "reason" > > and put in "rule." e.g. can a simple rule based system start an > > expanding system - expanding by virtue of it's acquiring facts and > > rules. More about this below. > > > > > > Reason-based reasoning is just a strategy in which analysis and > response > > > to a situation is based on reasons which the AGI program can access. > > > > rules no? > > > > > .... <section removed > > > > > > > So my question is whether or not reason-based reasoning can be used > > > effectively in a simplistic system to enable the program to make good > > > reactions based on what it had learned. But I do not fully understand > > > how human beings are able to adeptly recognize and react to complicated > > > situations. > > > > stan: Would it be fair to rephrase this question? Something like " can > > one build a simple rule based system that will make good choices based > > on rules and facts it has acquired?" > > > > If so, it seems obvious that the system can only operate on rules and > > facts acquired. > > > > The bigger question is can the system take a fact or rule gathered from > > it's experience in the environment and place it where it's rule > > processor finds it at the right time? What are the rules that tell the > > system how to utilize this 'mined' knowledge? > > > > simple huh? > > > > > > > > Analysis and reactions do not only act on some form of output. They can > > > govern the analysis and reaction modes as well. > > > > stan: True, the "output" could simply be a state change in the > > analyzer. This often is the case when writing a program - branch on a > > condition... > > > > ... portions removed > > > > > > One problem that I do not completely understand is how concepts are > > > integrated. Reason-based reasoning will help but it does not explain > > > everything. I am thinking about starting with a primitive artificial > > > language to make the program work a little like a programming method. > > > However, with reason-based reasoning that is able to act on recognition > > > and reaction methods there is no reason why I could not experiment with > > > language acquisition. > > > > > stan: If the goal is to understand how concepts are integrated, put on > > the designer hat and say "how would I design a system that can integrate > > concepts?" What do the players look like? > > "Concepts" is a bit nebulous to me. More concrete might be something > > like " how are rules and facts integrated in the system I am designing?" > > > > take care... > > Stan > > > > > > > > ------------------------------------------- > > AGI > > Archives: https://www.listbox.com/member/archive/303/=now > > RSS Feed: > https://www.listbox.com/member/archive/rss/303/19999924-5cfde295 > > Modify Your Subscription: https://www.listbox.com/member/?&; > > > Powered by Listbox: http://www.listbox.com > *AGI* | Archives <https://www.listbox.com/member/archive/303/=now> > <https://www.listbox.com/member/archive/rss/303/10561250-164650b2> | > Modify<https://www.listbox.com/member/?&;>Your Subscription > <http://www.listbox.com> > ------------------------------------------- AGI Archives: https://www.listbox.com/member/archive/303/=now RSS Feed: https://www.listbox.com/member/archive/rss/303/21088071-c97d2393 Modify Your Subscription: https://www.listbox.com/member/?member_id=21088071&id_secret=21088071-2484a968 Powered by Listbox: http://www.listbox.com
