Re. of(f) course

[christer@galaxy.ling.lu.se (Christer Johansson) (by way ofvanessa@media.mit.edu (Vanessa Stevens Colella))]

The collective pilot problem might have an interesting and complicating
factor: it seems like all the "pilots" are
running some kind of individual "selfish" simulations of what the outcomes of
their activities are.

  It might be that a decision system could possibly rely on the set of
reactions
as a (the?) set of solutions and select only from that.
  With feedback (as Bruce Gregory pointed out) this might avoid the average
behaviour. Can we trust each individual to give a valid solution? Maybe YES if
the individual is such a complicated unit as a trained human pilot.
  One remaining question might be timing - some "pilots" will have their
planning interupted, maybe they should be given some time to conform to the
emergent plan before being considered as advisors for that plan?

  /Christer Johansson