The board games question is interesting, because it highlights an implicit assumption of my approach to conceiving the generalizable/ creative goals/tests of AGI.
I'm thinking in robotic terms of **the basic, natural activities that more or less any agent/animal must undertake to survive in the real world **- such as navigating the terrains of the world, manipulating the objects of the world with effectors and/or bodies (as worms do), perceiving and categorising/classifying the objects of the world and their diverse classes and properties. All of these are general, creative activities in which the agent must navigate/manipulate/perceive one diverse group of objects after another. Board games are a highly evolved, abstracted version of these. Instead of directly navigating the world - from navigating forests to opponents in physical battles to opposing players on a field, we merely imagine ourselves doing so in a stylised way in, say, chess or draughts or shoot em up videogames.. So to start with board games is rather artificial - and won't I think, work or be fruitful, much as to start in an artificial virtual environment rather than a real one won't work. Artifficial activities actually depend on having embodied knowledge of real world, comparable activities, much as manipulating the numbers and abstract forms of maths depends on having knowledge of manipulating real world objects. But I await any response from Ben or others. ------------------------------------------- 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
