One consequence of dynamic system theory is that relatively simple systems, which can be described by a few non-linear equations, can exhibit very complicated, stochastic-like behavior. Such models simulate processes inexpensively. They reveal insights into the underlying mechanisms while devising strategies to control these processes.
Chaotic systems are sensitive to initial conditions. Since these conditions are not precisely known and are subject to perturbations, long-term predictions of the behavior of these systems are impossible. Thus, the availability of large computational resources will not enable one to generate long-term predictions for systems ranging from weather to economic forecasts.