Computer chess offers two opposite perspectives on measuring how humans are at making good decisions.
On one hand, these days computer programs stomp all humans, so it shows we often do not, cannot, make optimal decisions, no matter how hard we try. This will likely be the case if "ground truth" ever becomes available for any other human decision.
On the other hand, under certain circumstances, we can find the optimal action for fairly difficult problems. One example is the Troitzky line in the KNNKP endgame, discovered long before computers but subsequently confirmed correct.
I'm guessing an infrastructure of careful deliberation helps, with one person trying to refute another's published analysis.
(all with the backdrop of the political season)
No comments :
Post a Comment