In many animal species, individuals acquire knowledge from others that enhances their survival and reproduction. However, among the many available exemplars, not all provide reliable information. Consequently, individuals tend to choose their exemplars selectively. One widespread pattern is a preference for older individuals, who may have accumulated valuable knowledge over their lifetime. Yet empirical studies also show that individuals frequently learn from peers of similar age, suggesting that copying elders is not universally optimal. What ecological and social conditions favor learning from elders rather than peers, therefore, remains unclear.

Here, we investigate the evolutionary drivers of age-biased exemplar choice in age-structured populations where individuals accumulate knowledge over their lifespan. We show that the evolution of exemplar age choice is shaped by a trade-off between targeting knowledgeable individuals and accessible ones. This trade-off leads to learn preferentially from relatively young exemplars, who are common and still able to provide substantial amounts of novel knowledge, given learners’ limited knowledge at early ages. As individuals age, encountering exemplars with substantially novel knowledge becomes increasingly difficult. Consequently, as they get older, individuals are expected to shift toward learning from older individuals, who possess more knowledge. Population, environment, and knowledge characteristics can shift this balance, generating a wide range of strategies from learning primarily from peers to consistently targeting the oldest individuals.