Over the past decade, academics have developed a multitude of risk models specifically designed to capture time and space amplification. In tandem, we have seen an exponentially increased reliance on risk models by regulatory authorities, insurance and financial institutions, and government agencies. But have these new risk models resolved uncertainty? On the contrary. Their increased complexity and widening scope of application have induced a dramatic increase of model risk. The time and space amplification features are crucial but also entail that existing knowledge about risk evaluation is of little use when dealing with modern risk models. Hence, we need new approaches to conduct modern risk evaluation, that respond to the vast uncertainty of 21st-century risks.