>Here's a potentially better protocol:
>First the entity must ask itself, what am I trying to accomplish, what am I
>trying to become, what is my goal? Setting a goal is essential to actually
Agreed. Except, of course, there would most likely be a fairly large set of goals.
>Then the entity sets up a worldspace in which the entity populates with
>copies of itself, which can interact, combine, reproduce by both asexual
>and bisexual intercourse, as well as being able to edit themselves and each
>other on a limited basis.
>Entity copies within the worldspace can grow over time, but if they remain
>static they die after a given amount of time. Those that lose various
>competetive rankings more often than not die sooner rather than later.
>competetive rankings are based loosely on the goals the Entity Prime wishes
>attain. Over time, the entities in the world space evolve toward the goals
>the Entity Prime, to the point that those entities which attain the goals
>Entity Prime can exit the world space. At this time, the Entity Prime can
>overwrite itself with Entity(n).
How about I, Entity Prime (EP) maintain my prime identity and also generate
(let's say) 25 sub-identities whose purposes are to achieve a set of goals
set forth by me at the time I create them? They are virtual copies except
for the executive functions which only the (EP) holds. They may
interact and edit themselves but not each other, except through the rational
processes of logical and persuasive argumentation. They may form work
groups/committees to hasten their progress through cooperation. As time
passes, sub-identities develop diverse attributes and collect specialized
information. One problem I see, and Doug already addressed this, is that
EP may fall too far behind to recognize and appreciate a truly positive opportunity for change. I also doubt that it will be in the best interests of
the EP in terms of efficiency to simply integrate all the attributes and all the specialized information gathered from all of the sub-identities. So perhaps the best path toward achieving the stated goals (as well as determining whether those goals are still worthwhile) is to require a consensual
democratic decision on what information will be ultimately be fully integrated
into the EP. This way the EP can not be allowed to unduly retard it's own evolutionary growth, plus the idea that an evil sub-entity might "fool" the EP into being overwritten is no longer a problem. It's no longer a question of either the EP or the copy surviving. This kind of group cherry-picking protocol may be more controlled and stable, but might also allow for a steadier and more sustained growth.
This idea is probably an old one, though. I'm still trying to catch on so bear with me.