Setting jokes about "Neanderthal conservatives" with their black-and-white thinking aside, and setting aside more serious issues about whether or not Bayes' theorem is the best way to handle how we adjust our thinking to changed scenarios, there's growing evidence that our brains evolved to handle probabilities.
If this is true, there's other fallout.
Above all (sorry, Ray Kurzweil, other singularity touters, etc.) this is another major dent for those who say major advances in artificial intelligence are just around the corner. There's nothing to indicate that robots, or even Watson-like computers regularly engage in this type of thinking evaluation. At the least, as the story notes, robots don't have anything close to the sensory skills to do something like this, and Watson has no sensory inputs of that sort at all.