Current models of curiosity postulate that it motivates behavior towards the resolution of uncertainty. Yet in everyday life, we often prolong uncertainty, for example avoiding spoilers for just-released movies. We investigated whether higher curiosity favors delaying resolution to experience an event as it unfolds. We developed a video task in which line drawings slowly resolved into objects. As each video progressed, participants (N=1338, Prolific) made choices about how long to keep watching. We showed that when curious, people often choose to remain uncertain as information is gradually revealed, rather than alleviating uncertainty immediately (e.g., viewing a spoiler). Moreover, contrary to prevailing models, we demonstrate that higher curiosity induced enjoyment instead of discomfort. Moreover, choices to delay resolution enhanced satisfaction by and improved memory for drawings. Our findings support a reconceptualization of curiosity as a state that can steer us away from instant gratification and towards the patience that enables discovery.

To guide effective decision making in an uncertain world, humans must balance seeking relevant information with the costs of delaying choice. Optimal information sampling requires computationally expensive value estimations. Information sampling heuristics are computationally simple, but rigid. Efficient and flexible information sampling, therefore, must leverage the advantages offered from both approaches. In the present study, human participants completed an information sampling task, in which they sampled sequences of images (e.g. indoor and outdoor scenes) and attempted to infer the majority category (e.g. indoor or outdoor) under two reward conditions. We examined how behavior maps onto potential information sampling strategies. We found that choices were best described by a flexible function that lay between optimality and heuristics; integrating the magnitude of evidence favoring each category and the number of samples collected thus far. Integration of these criteria resulted in a trade-off between evidence and samples collected, in which the strength of evidence needed to stop sampling decreased linearly as the number of samples accumulated over the course of a trial. This non-optimal trade-off best accounted for choice behavior even under high reward contexts. Our results demonstrate that unlike the optimal strategy, humans are performing simple accumulations instead of computing expected values, and that unlike a simple heuristic strategy, humans are dynamically integrating multiple sources of information in lieu of using only one source. This evidence-by-costs tradeoff illustrates a computationally efficient strategy that balances competing motivations for accuracy and cost minimization.