Emergence and Diffusion of GPTs- A Complex Systems Perspective

General purpose technologies (GPTs) are transformative innovations such as the steam engine, electrical transformer, digital computing, or artificial intelligence that catalyze broad economic and societal changes. They not only revolutionize individual sectors but also trigger waves of complementary innovations and organizational transformations across the economy and society.

This paper adopts a complex systems perspective, highlighting the nonlinear dynamics and emergent feedbacks inherent in innovation ecosystems to examine how GPTs emerge, diffuse, and co-evolve with interdependent innovations and institutions.

We introduce and formalize the concept of general purpose systems (GPSs) as the infrastructural and institutional frameworks that co-develop alongside GPTs, enabling and amplifying their widespread impact. Building on causal models of technological diffusion, including logistic S-curve adoption dynamics, network externalities, economies of scale, path dependence, and reinforcing feedback loops, we illustrate how the synergistic interplay of these processes governs the trajectory of GPT deployment and uptake. A macrohistorical analysis highlights the role of successive GPT-GPS constellations in driving epochal transformations in production, organization, and society, as well as divergence in development trajectories across civilizations, from the Industrial Revolution to the digital age.

Drawing on these insights, we discuss design and policy implications for contemporary and future GPTs—such as artificial intelligence, the Internet of Things, and biotechnology, emphasizing the need to foster complementary systems, adapt institutions, and shape incentives to harness these technologies for inclusive and sustainable progress. By explaining the co-evolution of technological and institutional innovation through a complexity lens, our study provides an integrative framework and strategic guidance for innovators, policymakers, and scholars navigating the next generation of GPT-driven transformation