On July 6, Goldman Sachs emphasized that 'world models' could become the second engine driving future demand for AI infrastructure. Unlike large language models that primarily process text and images, world models aim to understand causal relationships within physical and social systems, such as simulating friction, material behavior, supply chain reactions, policy shocks, or competitive strategies. Physical world models will support robotics, logistics, autonomous driving, and industrial design; social world models may be used for strategic simulations, investment decisions, governance stress testing, and policy scenario analysis. Goldman Sachs believes that world models will not replace large language models but will add new computing demands. If their development pace exceeds expectations, current investment forecasts regarding computing power and electricity may still be too low.