A Novel Metric for Machine Decision-Making Ability (MDMA)

Delving into the historical evolution of computing, this paper presents a compelling argument that the autonomous momentary mechanical on-off button serves as a conceptual bridge, pointing back to ancient devices that mark the true origin of computers—a moment when machines first gained the capacity to independently change their internal state. Unlike symbolic tools like the abacus, which required continuous human manipulation, the on-off switch introduced a foundational leap: the ability to decide—to be “on” when previously “off” and vice versa—without further intervention. This binary autonomy represents the first seed of machine intelligence. To quantify such capability, we introduce a novel metric: Machine Decision-Making Ability (MDMA), which measures a system’s potential for autonomous state transitions. A basic on-off switch is assigned MDMA = 2, while early machines like Pascal’s calculator demonstrate higher, approximate MDMA scores based on their mechanical depth and logic, such as carry operations. We also analyze simpler, partial-autonomy systems like ancient alarm clocks and Zhang Heng’s seismoscope (MDMA = 1), which represent a single irreversible decision event. By reframing the history of computing through the physical act of autonomous switching—rather than symbolic computation—this paper uncovers a previously overlooked turning point in the rise of machine intelligence and lays the groundwork for a universal scale of computational autonomy.