opencog-dev team mailing list archive

["Joel Pitt" <joel.pitt@xxxxxxxxx>]

On Thu, May 29, 2008 at 9:37 PM, Bob Mottram <fuzzgun@xxxxxxxxx> wrote:
> Question: In order for the system to be able to create models of
> itself should MindAgents also inherit from Atom ?

I think in the long term, MindAgents or system processes will be
represented as executable schema that are themselves made up of atoms.
The MindAgent class is just a stop gap until we've reached the
computational resources and opencog has the system maturity to
implement them in this way.

I guess if MindAgents will eventually be atoms anyhow, it might make
sense to give them the functionality of atoms. Although I'm not sure
what, if any, breakages this would cause in the current
implementation.

J