Thanks Ben, Bruce, and others for organizing this. I think it's a very significant effort to consolidate AGI interests among researchers.
I don't have a detailed AGI theory yet, hope to make it in the next round! =)
"What (if any) practical applications do you think will be enabled by your system at an intermediary stage, before it has achieved human-level AGI?" -- My approach to AGI is to build a passive truth maintenance system, ignoring all aspects of behavior generation at first. This may be a way to reach a useful landmark earlier.
In Novamente nodes may contain procedures. IMO this makes the knowledge representation very complex. In my model I use a "flat" representation akin to predicate logic / semantic network. This is one of the key assumptions I make, ie that a flat representation is sufficient for AGI.
YKY
To unsubscribe, change your address, or temporarily deactivate your subscription, please go to http://v2.listbox.com/member/[EMAIL PROTECTED]
