In this talk, we will discuss a new framework for incorporating model uncertainty into decision-making under uncertainty. To develop a mathematical formalism for decision-making, one usually studies functionals on the space of random variables, where the interpretation of the functional depends on the problem's context. For example, functionals can be used to calculate capital requirements in quantitative risk management or to calculate premiums for insurable losses in actuarial science. In practice, one usually begins by estimating a probabilistic model and using law-invariant functionals. This approach is convenient because the calculation of law-invariant functionals is tractable and leverages the extensive literature on statistical model building. However, there is often considerable uncertainty about the probabilistic model estimated from the data. To address this model uncertainty, one often resorts to distributionally robust optimization. This is done by fixing a collection of competing probabilistic models and computing the worst-case value of the law-invariant functional under them. As we will see, distributionally robust optimization can be seen as an evaluate-then-aggregate (ETA) approach to model uncertainty. This will then motivate us to discuss an aggregate-then-evaluate (ATE) approach. After defining the ATE approach, we will discuss ways to ensure that the decisions made through it respect normatively appealing properties, e.g., the preference for diversification. Finally, we will conclude this talk by showing that the ATE approach better supports decision-making when the statistician is a Bayesian. To do this, we will show why the ETA approach is not the right way to go about decision-making under uncertainty as a Bayesian, and how the ATE approach does not run into the same issue.