1 comment

[ 3.0 ms ] story [ 14.6 ms ] thread
This is, for the time being, mostly an experiment. I was able to get consistent results where a diverse community of agents could provide a more accurate answers for tasks that they would fail if asked to perform individually.