What inherent limitation of rule-based architecture drove the conceptual shift in the 1990s?
It required extensive manual reprogramming when new medical guidelines emerged
The architecture utilized by early systems like MYCIN and Internist-I, characterized by extensive sets of hand-coded 'if-then' rules, presented a major practical hurdle as medical science rapidly advanced. Because knowledge engineers had to anticipate nearly every potential scenario, updating the system to reflect new diagnostic criteria or standard-of-care treatments—such as the introduction of a new drug protocol—necessitated time-consuming, manual reprogramming of the symbolic logic paths. This rigidity directly contrasted with the emerging philosophy of Evidence-Based Medicine (EBM), which demanded systems that could actively incorporate and synthesize the latest published research rather than relying solely on static, pre-programmed expert knowledge.
