Meaningful decision-making requires prediction. Without some rough, possibly vague, guess at the consequences of different available decisions, we would have no reason to prefer one decision over another. The important question is what type of prediction best supports decision-making.

We need to study the Kimberley system as a whole, addressing specifically the cumulative effects of various human uses and activities. This is a challenging task. There is a high level of uncertainty about how the Kimberley system works and what future stressors it may encounter. Uncertainty is not the only challenge. The future of the Kimberley system will be determined by the interaction between many forces: the economy, ecological processes, social development, climate change, human population dynamics, resource extraction and many others. Notice that these processes influence one another, leading to potentially very complex dynamics.

A vast literature agrees that humans, including very bright individuals and experts, can be very poor at predicting how systems behave when they undergo complex dynamics. In fact, most people fail even when they analyse embarrassing simple systems. Computer models help us reducing the extent of these mistakes. In general, model predictions are more reliable than the ones we (expert included) may produce without models. Of course, models will not accurately predict what will happen to the system. Rather, they are designed to say something physically and ecologically meaningful about what may happen to the system, should it undergo specific pressures.

Our models are not microcosms. They are neither virtual laboratories nor digital representations of the immense complexity and richness of the Kimberly region. They are tools which allow us to integrate the available knowledge, include the state-of-the-art understanding of social, economic and ecological processes and account for uncertainty and missing information. We can use them to explore the possible consequence of different human decisions and actions, circumventing many of the well known limitations of the human brain at handling complex dynamics and information. They help us understanding the system and make better decision.