Hi Ben, I've been looking at the same problem from a different angle... rather than searching for simplified artificial worlds for an agent to live in, I've been searching for models of the world to be used directly for reasoning (i.e., the internal world for an agent situated in the real world).
I'll be interested to see what you do with CogDevWorld, however you may be interested in some of the specifics of my approach: I take the approach that *everything* should be "beads". I model the world with 'beads' (that are point masses, and carry heat, color, etc), 'joins' to structure them together (and that can change under stress or heat) and 'surfaces' (convex hulls around small set of beads) that stop things from being permeable. Solid objects can be constructed from beads with rigid (but 'snap'-able) joins; flexible objects by beads with flexible joins; gases by weightless unjoined beads, and liquids/adhesives by 'local joins' that attract and repel their neighbours (to create incompressible liquids with surface tension). I like this approach because everything is uniform - not only does the same mechanism simulate liquids, solids and gases; but more importantly, new laws of physics can be added 'orthogonally' to the existing laws. The approach is flexible and open-ended. I haven't had the resources or time to explore the different laws of physics for such models, but it isn't hard to create realistic objects in simple environments. This approach is far more computationally expensive than, say, rigid-body physics of complete object models; but the locality of the computations means that it should scale very well. (In fact, the orthogonality of the laws of physics means that you could support many more physical laws than could be reasonable computed simultaneously, but only enable laws as they are relevant to particular problems - e.g., turning on laws of heat diffusion and state-of-matter changes only when the agent is interacting with, say, the fridge or fire). I outlined the basic principle in this paper: http://www.comirit.com/papers/commonsense07.pdf Since then, I've changed some of the details a bit (some were described in my AGI-08 paper), added convex hulls and experimented with more laws of physics; but the basic idea has stayed the same. -Ben -----Original Message----- From: Ben Goertzel [mailto:[email protected]] Sent: Saturday, 10 January 2009 9:58 AM To: [email protected] Subject: [agi] What Must a World Be That a Humanlike Intelligence May Develop In It? Hi all, I intend to submit the following paper to JAGI shortly, but I figured I'd run it past you folks on this list first, and incorporate any useful feedback into the draft I submit This is an attempt to articulate a virtual world infrastructure that will be adequate for the development of human-level AGI http://www.goertzel.org/papers/BlocksNBeadsWorld.pdf Most of the paper is taken up by conceptual and requirements issues, but at the end specific world-design proposals are made. This complements my earlier paper on AGI Preschool. It attempts to define what kind of underlying virtual world infrastructure an effective AGI preschool would minimally require. thx Ben G -- Ben Goertzel, PhD CEO, Novamente LLC and Biomind LLC Director of Research, SIAI [email protected] "I intend to live forever, or die trying." -- Groucho Marx ------------------------------------------- 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/?&; Powered by Listbox: http://www.listbox.com ------------------------------------------- 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=126863270-d7b0b0 Powered by Listbox: http://www.listbox.com
