Judgment under radical uncertainty: Epistemic rational heuristics

Abstract

The special issue “Expanding the boundaries of rationality: Towards new models of decision making for radical uncertainty” addresses the challenge of developing logically sound models of decision making for conditions of uncertainty evading those manageable by established decision making approaches. This editorial characterizes the sought new models and approach as constructivist in nature but neither deductive nor Bayesian in logic: a set of epistemic rationality models for discovering new problems and solutions (“found” either in the real or the artificial world, i.e., either detected or designed) that extend the available bounded rationality models. This editorial offers a map of this territory, positioning the new heuristics identified in this collection of articles in a more general picture of heuristic methods qualifyable as fit to radical uncertainty.

The quest for non‐Bayesian decision‐making heuristics: Towards a logic of imagination

“…Judgment may also require “imagination” (see also Foss, 2023, and Grandori, 2023, this SI). This idea was already in Simon (1978) when he discussed “creative” problem solving and design.…”

“…Likewise, a scientific model is used to detect anomalies and organize the minimal revision that fits with the surprising data. Thus, scientific heuristics (hypothesis, experimentation, and modelling) that aim at detecting new causes and anomalies (see Grandori, 2023, this SI) are possible guides for reflexive judgment and for navigating in the unknown.…”

Artificial intelligence and radical uncertainty

“…We will next discuss in more detail how AI can be used in different decision‐making tasks as a ‘heuristic’ to help address radical uncertainty. In particular, we shall focus on the role of AI in situations of ‘unknown‐unknowns’ or decision‐maker ignorance (see, e.g., Foss, forthcoming; Grandori, 2023 and Hatchuel, forthcoming, this Special Issue). Current work on AI emphasizes its generative elements, or its ability to augment human knowledge (e.g., Eapen et al, 2023), which fits well to our need to in management and organization studies to better understand decision‐making in situations where humans are ignorant of critically important contingencies and alternatives.…”