PM, You might be good on this. I've been realising recently that one of the keys to understanding language/.a-conceptual/system is to recognize that it conceptualises hierarchies of action.Arguably, that is its central - even its only function.
Thus any complex course of action, like GO TO THE KITCHEN can be broken down into subcourses of action on several levels, such as WALK TO THE DOOR, OPEN THE DOOR, TAKE A STEP etc. This applies to the whole of language - OBAMA AND NETANYAHU HAD DIFFICULT DISCUSSIONS can be broken down into an elaborate hierarchy of levels of action. Similarly, and even more so, general abstract statements like NEGOTIATIONS BETWEEN STATE LEADERS ARE OFTEN DIFFICULT Who are the main theorists on this area? [I've just remembered one - James Albus] But there must be more, though I suspect this area is so computationally/robotically obvious, that it hasn't been explored remotely enough. One of the fascinating and central (including central to AGI) aspects of any real-world hierarchy of action, is that it is necessarily both open-ended and "open-levelled". I doubt that anyone will have picked that up On 6 November 2013 14:50, Piaget Modeler <[email protected]> wrote: > James Allen also did research in this field. His book *Reasoning About > Plans* > should be pretty informative. > > ~PM > ------------------------------ > From: [email protected] > To: [email protected] > Subject: RE: [agi] "Singularity or Bust" film is now online for free ... > Date: Fri, 1 Nov 2013 17:29:42 -0700 > > --------------------------------------- > > > > > There was also mention of "the stories we tell ourselves about our own > > motives", and that's something which I was considering recently. Given > > a set of events stored in an episodic buffer how is it possible to > > figure out the "why" part of each circumstance? Being able to infer > > motives, both in others and also in yourself seems to depend upon being > > a member of a storytelling culture in which particular characters in > > mythical or allegedly "real" circumstances (such as news articles) are > > ascribed certain motivations. > > > > Some amount of motive inference would also be possible without being > > part of a culture, where the motives are really just basic drives such > > as eating. > > > > --------------------------------------- > > From my understanding the "why" or motive requires having a model of > actors (roles), plans, and goals. Once you have this model, then observed > events and episodes can be interpreted against the model. That is what > most of Roger Schank et. al.'s work was about. Building such a model and > then using it for story understanding and episode and motive explanation. > > Matching episodes and events to actors, plans,and goals will give you > motives. You just need some good algorithms to (a) bulid your model, > (b) accrue your events and episodes, (c) match episodes and events > to the mental model. > > Bob's your uncle. The devil is in the details though. Just my opinion. > > ~PM > > > *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 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
