[ExI] ants again

Alejandro Dubrovsky alito at organicrobot.com
Sun Apr 5 10:20:16 UTC 2009


On Sat, 2009-04-04 at 15:00 -0700, spike wrote:

>  
> As a preliminary step to the double straw, we could set up the double
> hole experiment in which the ants are allowed to set up their aphid
> farms in the tree using a paper bridge over the goo.  The bridge
> has goo barriers with two adjacent holes large enough for one ant to
> pass easily but not large enough for two to squeeze by each other.
> The holes would be about two mm diameter and about four mm center to
> center.  The experiment is to see if the ants can figure out a way to
> make one hole the in and the other one the out.  The holes are
> circumscribed by a goo ring so the ants cannot go around the holes.
> It's thru the holes or no access to the aphids.
>  
> Anyone wish to speculate on whether the ants will split their lines
> and form two one-way holes, thereby reaching their aphids relatively
> impeded, or whether they will get in each other's way in both holes.
> Do state your reasoning.

I know little of the organic variety of ants, but if I was to set up a
run of ant colony optimisation
(http://en.wikipedia.org/wiki/Ant_colony_optimization) with those
parameters, I'd be quite dissapointed if the virtual version of the dumb
beasts didn't find the optimised path.  Are the stochastic parameters in
the organic variety really that low that you wouldn't get enough of them
going through the second straw often enough to cause the stabilisation
of the second path?  What about the chemical signals they use, don't
they diffuse even those measly 2mm?  If neither, how long would it take
you to evolve those organic robots to have the creative capabilities of
their virtual brethren?





More information about the extropy-chat mailing list