① As generative AI rapidly diffuses across the economy, expectations for a productivity revolution are growing, yet macroeconomic productivity indicators have not yet shown clear improvement. Using a household survey, this paper empirically examines whether AI adoption generates potential productivity gains through reductions in work time, and whether such time savings translate into realized output growth.

② The analysis finds that AI adoption reduces average work time by 3.8% (approximately 1.5 hours per week). This effect is particularly pronounced among lower-skilled workers and heavy AI users. If we assume the time savings are fully converted into productivity gains, the estimated potential productivity improvement is approximately 1.0%.

③ However, these time savings do not translate into actual output growth (essentially zero correlation). While AI has improved efficiency at the individual task level, it has not extended to workflow improvement, organizational restructuring, or labor reallocation, resulting in a "productivity disconnect." As an exception, productivity gains were observed among the self-employed, professionals, and intensive AI users — groups with strong performance incentives and high job autonomy — suggesting the role of organizational structure and incentive systems in determining AI’s effects.

④ AI has currently entered the 'efficiency' stage but has not yet fully transitioned to the 'productivity' stage. This can be viewed as a typical transitional process (J-curve, Solow Paradox) in the early phase of general-purpose technology adoption. Future productivity paths will vary depending on policy responses and the transformation of corporate organization and labor market structures. Realizing AI's productivity effects requires redesigning work processes and organizational structures (standardized tasks vs. open tasks), job reallocation, and building performance-based incentive systems. Continuous monitoring of shifts in young workers' skill formation pathways is also needed.