Advice-taking plays a critical role in collaboration. Yet people tend to under-utilize advice, often to their own detriment. We investigate if people's utilization of advice improves when they know the advisor has access to different information compared to them. We examine how individuals integrate advice in an estimation task, where the advisee and the advisor have access to different perspectives of the same problem. We assess how individuals adjust their estimates when presented with estimates from a human advisor and an AI advisor, and when they are given information about the advisor's perspective. Our findings are consistent with egocentric discounting where individuals exhibit a general bias toward their own information. However, this discounting is lower for AI advisors compared to human advisors in our experiment. Our results also show that the advisor's estimate is taken more into account when the advisor has a more favorable viewpoint -- for both human and AI advisors. This suggests a potential for optimizing advice-taking behavior by enhancing people's understanding of the advisor's viewpoint. This study furthers our understanding of advice-taking dynamics with human and AI advisors and the role of perspective-taking in decision-making processes.