Exploring by Believing
Abstract: Sometimes, we face choices between actions most likely to lead to valuable outcomes, and actions which put us in a better position to learn. These choices exemplify what is called the exploration/exploitation trade-off. In computer science and statistics, this trade-off has fruitfully been applied to modulating the way agents make choices over time. In this talk, I'll argue that the trade-off also extends to belief. We can be torn between two ways of believing, one of which is expected to be more accurate, whereas the other looks like it will lead to more learning opportunity. Further, it is sometimes rationally permissible to choose the latter. I break down the features of action which give rise to the trade-off, and then argue that each feature applies equally well to belief. This result hangs on the connection between what we believe and how we imagine. I'll end by presenting some preliminary experimental work testing whether humans actually do believe in an exploratory way.