On Sat, Dec 15, 2012 at 1:01 PM, Mike Tintner <[email protected]>wrote:
> Logan:. The representation needs to support activation. > > imperative statements, like do this or that! > > Just trying to understand you here. Do you mean by “activation” - > “putting into action” ? > yes > > So “go to the kitchen” , you would say, “activates” the course of > action of going to the kitchen (pace the Woz test) ? Right? > actually it would be "do go to the kitchen" that activates or is imperative, wheras "ya go to the kitchen" is declarative or passive. > > How then do you see the *internal representation* of words, such as “go to > the kitchen,” translating into external courses of action? > You could have some internal programs or stories, that would be interpreted in order to perform actions. So for instance going could be mean multiple things, though a simple story or program of getting there would be. the go story get your current position. get your destination position. plot a course from current to destination. if short distance then walk is method. if moderate distance then bike is method. if long distance then transport is method. use method from current position to destination position. recognize arrival at destination. story end If for instance this robot AGI was mobile and powered by some fuel food that it kept or prepared in kitchen, when it ran low, it would look at what it's options are, one of which could be to go to kitchen. If it chose to go to kitchen then it could tell itself "do go to kitchen" and it would go through the go story to get there. > > > > > *From:* Logan Streondj <[email protected]> > *Sent:* Saturday, December 15, 2012 5:01 PM > *To:* AGI <[email protected]> > *Subject:* Re: [agi] Internal Representation > > well I heard more recent research actually pegged short-term memory around > 4, and that's the same as there are working registers in intel assembly , > ax, bx, cx, dx. > > Internal representation can be done like we humans do it, with human > language. > > quo > 1. The representation needs to support activation. > quo te > > imperative statements, like do this or that! > > quo > 2. The representation needs to support relationships (patterns among > elements). > quo te > you be similar to so and so by these features. > > quo > 3. The representation needs to support reification. > quo te > Yes, programs can be treated as data, > just as sentences can be instructions or information. > > quo > As long as the representation does that, I'm satisfied. > quo te > I'm glad you agree :-) > > > On Fri, Dec 14, 2012 at 5:42 PM, Piaget Modeler <[email protected] > > wrote: > >> >> Insightful idea. >> >> ------------------------------ >> Date: Fri, 14 Dec 2012 11:05:43 -0600 >> >> Subject: Re: [agi] Internal Representation >> From: [email protected] >> To: [email protected] >> >> Working memory, anyone? Maybe the fact that working memory is limited to >> about 7 items at a time isn't a weakness in the design of the brain. Maybe >> it's a feature with the purpose of limiting complexity. >> >> >> >> *AGI* | Archives <https://www.listbox.com/member/archive/303/=now> >> <https://www.listbox.com/member/archive/rss/303/5037279-a88c7a6d> | >> 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/303/6952829-59a2eca5> | > 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/303/5037279-a88c7a6d> | > 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-f452e424 Modify Your Subscription: https://www.listbox.com/member/?member_id=21088071&id_secret=21088071-58d57657 Powered by Listbox: http://www.listbox.com
