Chimera heuristics: Generative rational heuristics for the unknown from design theory

Abstract

The learning strategies offered by science for discovering the world by generating and testing hypotheses have been used abundantly to build decision-making heuristics. In contrast, decision-making heuristics for (re)designing the world are rarer. This paper develops a heuristic combining the exploratory power of chimeras with a design logic. Chimeras have long been used to foster imagination and build initially unknown futures. And recent advances in design theory show that in decision-making situations, chimeras can be generated as nonfalsifiable existential statements about desirable alternatives and events. Moreover, design theory offers learning operations that handle nonfalsifiable statements to generate new real objects. This paper uses these operations to build a rational heuristic that may or may not transform initial chimeras into reality. Its main effect is to ensure that stimulated learning leads to decision alternatives (whether pre-existing or novel) that surpass the initial optimal one. This paves the way for a class of design-based heuristics extending the main functions of Bayesian learning to a non-Bayesian world.