Creativity and Markov Decision Processes
Joonas Lahikainen, Nadia M. Ady, Christian Guckelsberger
TL;DR
This work addresses the problem of grounded creativity assessment for AI by formalizing mappings from Boden's process theory of creativity to Markov Decision Processes (MDPs) using the Creative Systems Framework (CSF). It introduces an abductive mapping strategy, analyzes eleven candidate mappings, and deeply evaluates three representative mappings (Ms, Msas, Mtau) to reveal how exploratory and transformational creativity, aberrations, and uninspiration could manifest in MDP-driven agents. The authors propose quality criteria to compare mappings and discuss limitations, with future work including more mappings, meta-level formalization, hyperparameter tuning, and extending to POMDPs. The overarching goal is to provide a theory-grounded, transferable toolkit for analyzing creativity across AI systems and domains.
Abstract
Creativity is already regularly attributed to AI systems outside specialised computational creativity (CC) communities. However, the evaluation of creativity in AI at large typically lacks grounding in creativity theory, which can promote inappropriate attributions and limit the analysis of creative behaviour. While CC researchers have translated psychological theory into formal models, the value of these models is limited by a gap to common AI frameworks. To mitigate this limitation, we identify formal mappings between Boden's process theory of creativity and Markov Decision Processes (MDPs), using the Creative Systems Framework as a stepping stone. We study three out of eleven mappings in detail to understand which types of creative processes, opportunities for (aberrations), and threats to creativity (uninspiration) could be observed in an MDP. We conclude by discussing quality criteria for the selection of such mappings for future work and applications.