Interesting: at the bottom of the article is his proof of “no-free will”. Apparently most people would get between 70-80% predictable. That doesn’t surprise me These are my results after 5 minutes: It did a worse and worse job at predicting me, but I suspect being musical and pattern averse generally, perhaps it may be the case that some people have more free will than others or rather operate less computationally. Still, predicting “what common people will do” shouldn’t be that difficult, especially if one is staying within a particular subculture or culture (such as the university testings he’s done where he generated 70%-80% results).
Still though, for a simple program, not bad. I’m sure it successfully spooked more than a few.
http://people.ischool.berkeley.edu/~nick/aaronson-oracle/
