Or Watson learns (through gradient descent of course) that you can make wild campaign promises to win the election and not actually execute on that policy because it's more beneficial to make grand promises while running and not rock the boat while in office.