CONCEPT
The Problem-Solving Model of Progress
Laudan's operational replacement for truth-based evaluation: a theory or tradition is progressive when it solves more problems and generates fewer anomalies than its competitors, assessed comparatively and revised continuously.
The problem-solving model is the analytical heart of Laudan's framework. It replaces the unanswerable question "Which theory is true?" with the tractable question "Which theory solves more of the problems we face?" Progress is measured as the increase in solved problems relative to the cost of the anomalies incurred. The measurement is comparative — no theory or tradition is progressive or degenerative in isolation, only in comparison to its alternatives. The measurement is revisable — a tradition that appears progressive today may become degenerative tomorrow if its anomalies accumulate faster than its solutions. The measurement is operational — it requires specifying which problems are being solved, which anomalies are being generated, and what evidence would alter the evaluation. Applied to AI, the model demands that both the triumphalist and elegist traditions specify their problem sets and be evaluated on the full set rather than the curated subset that makes each look successful.
In The You On AI Field Guide
Laudan's