Define "rational". Psychologists like Kahneman often say they're using some agreed-upon definition that exists outside their field. But applications of Bernoull

Define "rational". Psychologists like Kahneman often say they're using some agreed-upon definition that exists outside their field. But applications of Bernoulli and Bayes are an area of ongoing research, even narrowly [1][2]. More generally, no theory of general intelligence exists, and to assume you can trick intelligent agents like humans with simplistic experiments is unconvincing IMO. [1] http://en.wikipedia.org/wiki/Proebsting%27s_paradox http://en.wikipedia.org/wiki/Proebsting%27s_paradox [2] http://www.science20.com/hammock_physicist/statistical_physics_attacks_st_petersburg_paradox_resolved-96549 http://www.science20.com/hammock_physicist/statistical_physi...