Persistent CAs
- A problem we had in CABot3 was that CAs didn't persist
like psychological memories.
- We needed that for parsing, so we, in essence, programmed
the topology to fade away as needed.
- Outside of that, I did some stuff with small world topologies,
and with these, the CAs seemed to persist for the right lengths.
- In NEAL, we'll make those MT areas contain bits of small world
topologies.
- We'll try to get those objects to persist for psychologically
reasonable times once activated.
- This should occur even if the objects go out of view.
