Now
we have organisms in our computers digitally evolving to be a specialist.
Division of labor is usually seen as efficient. We laud creatures from insects
to captains of industry for operating highly differentiated systems. However it
is difficult to start at time zero and run experiments with people or insects.
As
a result the computer has come to our rescue once again and we will see if Adam
Smith and evolution give us the same end point. Goldsby et al in the Proceedings of the National Academy for Science have
given over their computer as a safe haven for digital organisms to evolve in
peace, free from predators whether free market or hungry critters (though
perhaps these two are the same)(1).
The
starting points are colonies whose members can send or receive messages so the
cost of changing to a different task is known. This knowledge isn’t lost, but
gets established in the genome so that subsequent generations can build on
decisions made by previous generations. There were 7 discrete logic tasks to
get on with. Different experiments put moderate or high cost on task switching.
The
control had no cost for switching so an organism could do whatever was needed
in its location and could thus evolve into a generalist colony with everybody
tackling everything whenever it came along.
The
cost of task switching was written in as a delay of so many turns before the
organism could start its new task. These delays were bad news for the colony
that would be slower to prosper and the colony had to decide whether the task
urgency outweighed the delay, and then store its knowledge in its genome for
the benefit of future generations. A high cost meant waiting 50 turns, while
the moderate cost meant waiting 25 turns.
Of
course the inevitable happened – evolution towards specialist colonies with a
strong division of labor. The higher the cost, the faster the evolution
occurred. Communication was important in that it enabled synergies to work
where groups managed to be more successful than individuals stuck without good
communication.
Looks
good for Adam Smith, but the problem comes in determining costs of switching in
the broader sense. It’s not just the cost to the colony, or worse the colony
owner, but the cost to the individual is forgotten. A termite, an ant or a bee
doesn’t have quite the same desires and ambitions as an assembly worker in a
car plant.
