The problem of induction is a philosophical concept illustrating why empirical observations can never definitively prove a universal rule, regardless of how many confirming instances are found. For example, observing thousands of white swans does not guarantee the truth of the statement 'All swans are white,' because the future discovery of a single black swan would instantly disprove the rule. This limitation explains why scientists consider even the most robust theories to be provisional and subject to revision.