Decision Making under Deep Uncertainty

Deep uncertainty describes the simple fact that we often do not know what the future holds. When decision makers assume otherwise, they risk overconfidence, missed opportunities, and pursuing policies brittle to surprise.

Deep uncertainty exists when experts or stakeholders do not know or cannot agree on:

1. appropriate conceptual models that describe relationships among key driving forces in a system
2. the probability distributions used to represent uncertainty about key variables and parameters, and/or
3. how to weigh and value desirable alternative outcomes.

DMDU provides concepts, tools, and multi-objective, multi-scenario decision support methods designed to inform and improve decisions that face such conditions. While DMDU methods are varied, they all emphasize considering a wide range of plausible futures; seeking policies which are “robust” over these futures rather than optimal for any best-estimate; and explicitly designing policies and implementation approaches to adjust over time in response to new information. Robust Decision Making (RDM) is a key DMDU methodology.

Traditional “predict-then-act” methods demand that we predict the future to act upon it. This mindset contributes to hubris and myopic overconfidence among experts, excludes voices from the conversation, fosters distrust among the general public, and distracts attention from the main task at hand – using the best available science and evidence to help decision makers fashion creative solutions that enable a diverse society to pursue its common and sometimes conflicting goals in the face of transformational change.

DMDU employs new technology in the service of classic ideas. DMDU uses computers to help people explore multiple pathways into the future, to stress-test proposed policies to identify their strengths and weaknesses from multiple points of view, and to identify policy-relevant scenarios. DMDU practitioners help stakeholders identify new, more robust strategies that meet multiple objectives over a wide range of futures, and help decision makers implement such strategies.

Rather that tell people what to do, DMDU is designed to help people work together to build a common understanding of challenges and to find creative compromise strategies. Its focus on human-machine collaboration aims to provide a “prosthesis for the imagination” and may offer avenues for empowering uses of AI. DMDU enables experts to build trust by being honest about uncertainty, and makes uncertainty empowering rather than something to be feared.

DMDU can restore trust in science by allowing experts to be humble in what they know yet confident in their recommendations.

Related Publications