Well I think I agree with you to the extent that you recognize that virtual reality can act as a kind of embodiment and that the features that are significant to AG in such a system would be the potential for interrelated complexity. But, with an advanced AGI capable of genuine learning, the same kind of potential for interrelated complexity that could enable progressive learning in virtual reality could be provided by teacher using some other system (like text-based) IO. In other words, information about the simulation of place (in your example) could be input through text.
The real question is whether or not advanced learning is feasible. Using virtual reality you have the advantage of pegging into a complicated database wherever you wish. Any feature your AGI program can't hack can be subsequently provided to it by the programmer. Eventually, your VR-AGI program would either get so complicated that you'd give up because it failed to demonstrate the kind of higher reasoning that was once wished no matter how complicated it gets, or you might discover that the secret has to be learned by the programmer before it can 'emerge' from some virtual reality. My idea of an AGI has it interacting with the IO data environment, such as a text-based IO, but it would have to be able to figure somethings out for itself right from the start. I can't describe what that means exactly, because the idea of a computer program itself is so closely related to the idea of categorical thinking that whenever I try to choose some words to define it, my choice of words tend to be ambiguous and need to be subsequently over- qualified. So maybe it is just a question of the degree and range of it's ability to figure some things out for itself. Whenever a programmer starts out with a highly detailed database and some preexisting paradigm as a base for his program, he can find that he can acquire some leverage by simplifying his program relative to the paradigm and the database. It is this kind of leveraged short-term gain that prevents the programmer from doing the hard work of discovering how to improve the program's learning abilities. Jim Bromer On Wed, Aug 13, 2008 at 12:02 PM, James Ratcliff <[EMAIL PROTECTED]> wrote: > How do you interact with, and how does your idea of advanced AGI act without > being embodied? > > I dont see how it can without being embodied (VR at least), *to some > extent*, and the more embodied, the more complicated actions that can be > allowed. > > A personal assistant AGI could be attached to a cell phone with GPS, and use > a VR of our real world to know where we are at all times, assist with > scheduling, and use google maps to plot out where we need to go and when, > but would have to internally represent all this data. > > What level of internal representation such as this will qualify as > embodiment? > > _______________________________________ > James Ratcliff - http://falazar.com > Looking for something... > > --- On Wed, 8/13/08, Jim Bromer <[EMAIL PROTECTED]> wrote: > > From: Jim Bromer <[EMAIL PROTECTED]> > Subject: Re: [agi] The Necessity of Embodiment > To: [email protected] > Date: Wednesday, August 13, 2008, 7:04 AM > > There is a another reason why embodied agi is useful. That is because > the challenge will provide some discipline for the programmer who > might otherwise never confront the structural problems that I believe > are fundamental to the problem of developing genuine agi. > Jim Bromer > > > ------------------------------------------- > 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 | Modify Your Subscription ------------------------------------------- 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=111637683-c8fa51 Powered by Listbox: http://www.listbox.com
