Little is known about the extent to which individual differences guide social learning. Here we use individual differences in intelligence and personality, specifically openness to experience, to predict the social learning strategies people use. Participants (N = 220) completed a range of general intelligence tests, a personality questionnaire, and a battery of novel behavioral tasks. In each behavioral task participants were trained on the solution to a problem. They were then informed of an alternative solution to the same problem that varied in terms of its quality, simulating new social information. The extent to which participants switched to a superior solution measured content-biased social learning, and the extent to which they retained the trained solution, when presented with an alternative solution of equal or higher quality, measured egocentric bias. As predicted by cultural evolutionary theory, most participants exhibited content-biased social learning. However, a significant minority (20-40%) exhibited an egocentric bias, preferring to retain the familiar, but often suboptimal, trained solution. Higher general intelligence was associated with general solution switching but was more strongly related to switching to a superior solution. So, higher general intelligence predicted content-biased social learning. By contrast, higher openness to experience, and therefore lower egocentric bias, was uniquely associated with switching to an inferior solution. So, egocentric bias (or lower openness to experience) was adaptive as it inhibited participants from switching to a maladaptive solution. Our findings highlight the importance of individual differences in intelligence and personality to human social learning.