Law of Accelerating Returns

The Law of Accelerating Returns is Ray Kurzweil's formalization of the pattern he observed across a century of computing history: that information technologies improve at exponential rates, that the exponential itself accelerates over time, and that this acceleration is driven by a specific mechanism—each generation of technology creates more powerful tools for designing the next generation. The law extends beyond Moore's Law, which describes transistor density on integrated circuits, to encompass all forms of information processing across five complete paradigm shifts. Kurzweil's claim is that the pattern is not coincidental but structural: it reflects the fundamental dynamics of how information-processing systems evolve when they possess the capacity to improve themselves. The law predicts not merely that technologies will improve, but that the pace of improvement will increase, producing a knee where steady progress becomes overwhelming transformation.