Multi-agent systems are widely used in implementing systems with autonomous behaviour, as well as simulating models that involve human behaviour. Ethical decision-making is a unique aspect of such socio-cognitive systems. In situations that require deliberation over a multitude of options with ethical implications, moral dilemmas are often resolved by a range of heuristics and principles that view the situation from different perspectives. While constraint and utility-driven decision-making strategies are customary in agent-based models, the incompatibility among multiple perspectives can invalidate the underlying premises of models of probabilistic utility-based decisions that rely on classic Kolmogorov axioms. In this paper, we highlight the fallacies of such models and present a practical case study to demonstrate how quantum cognition models can provide an alternative and credible representation of decision-making behaviour in simulations that involve ethical dilemmas.
Multi-agent systems are widely used in implementing systems with autonomous behaviour, as well as simulating models that involve human behaviour. Ethical decision-making is a unique aspect of such socio-cognitive systems. In situations that require deliberation over a multitude of options with ethical implications, moral dilemmas are often resolved by a range of heuristics and principles that view the situation from different perspectives. While constraint and utility-driven decision-making strategies are customary in agent-based models, the incompatibility among multiple perspectives can invalidate the underlying premises of models of probabilistic utility-based decisions that rely on classic Kolmogorov axioms. In this paper, we highlight the fallacies of such models and present a practical case study to demonstrate how quantum cognition models can provide an alternative and credible representation of decision-making behaviour in simulations that involve ethical dilemmas.
The present thesis (mostly) concerns the application of alternative mathematical methods to understand patterns in human cognition and to model them. The different chapters presented in this thesis show research that concerns the application of quantum probability (QP) theory in the modeling of human decision-making. Quantum probability (QP) theory is a theory for how to assign probabilities to events. QP theory can be thought of as the probability rules from quantum mechanics, without any of the physics. This work is not about the application of quantum physics to brain physiology. Rather, we are interested in QP theory as a mathematical framework for cognitive modelling. This theory is potentially relevant in any behavioural situation that involves uncertainty. QP theory is analogous to classical probability theory, though QP theory and classical probability (CP) theory are founded from different sets of axioms (the Kolmogorov and Dirac/von Neumann axioms respectively) and so are subject to alternative constraints. In this thesis we show that especially over the last decade, there has been a growing interest in decision-making and cognitive models using a quantum probabilistic (QP) framework. We see how this development encompasses publications in major journals (see Pothos and Busemeyer, 2013; Wang et al., 2014; and Yearsley and Pothos, 2014; among others), special issues, and dedicated workshops, as well as several comprehensive books (Busemeyer and Bruza, 2012; Khrennikov, 2010; and Haven and Khrennikov, 2010). However, uncertainty itself is neither ethical nor unethical – yet it is inherent to most situations in which, for instance, moral judgments and decisions have to be made. For a descriptive understanding of judgment and decisions in moral situations, it is an important lesson to acknowledge both the cognitive side (bounded rationality) and the environment (ecological rationality) – and thus the uncertainty of the world and how the mind deals with it. This thesis also shows significant interest in moral and social psychology. Specifically, we consider present technologies that suggest a need for evaluating alternative contexts for ethical decision-making. How the research on human-machine interaction feeds back into humans’ understanding of themselves as moral agents? This key question ultimately relates to the nature of ethical theory itself. Overall, this dissertation presents and addresses not only standard aspects of decision- making processes, such as similarity judgments (Chapters 1 to 4) or the constructive role of articulating impressions (Chapter 5), but also standard aspects of social psychology, such as moral judgments (Chapters 6 and 7) and game theory (Chapter 8). As stated in the Declarations section, the present thesis is a combination of a standard and a publication- based dissertation.
Artificial intelligence has long sought to mirror human intelligence through computational models of perception, learning, and reasoning. In parallel, cognitive science has shown that human judgment frequently deviates from classical logic and probability theory. This research-in-progress investigates how quantum-probabilistic models can be implemented as quantum circuits and examines the stability of such circuits, using the conjunction fallacy as a test case. Responses are compared across human behavioural data, Monte Carlo simulations, quantum simulations, and physical quantum hardware, explicitly testing how interference-driven reasoning patterns persist under controlled stochastic and decoherence-induced noise. This comparison provides insight into how cognitive probability structures are preserved—or distorted—across classical and quantum computational implementations. The paper demonstrates a proof-of-concept pipeline for encoding cognitively motivated probability structures in computational models and evaluating their behaviour under simulation and hardware noise, with implications for future decision support system design.
We consider the following model of decision-making by cognitive systems. We present an algorithm -- quantum-like representation algorithm (QLRA) -- which provides a possibility to represent probabilistic data of any origin by complex probability amplitudes. Our conjecture is that cognitive systems developed the ability to use QLRA. They operate with complex probability amplitudes, mental wave functions. Since the mathematical formalism of QM describes as well (under some generalization) processing of such quantum-like (QL) mental states, the conventional quantum decision-making scheme can be used by the brain. We consider a modification of this scheme to describe decision-making in the presence of two ``incompatible'' mental variables. Such a QL decision-making can be used in situations like Prisoners Dilemma (PD) as well as others corresponding to so called disjunction effect in psychology and cognitive science.
Cognitive Science is the study of the mind and how it relates to human behaviour. As a field, it is highly interdisciplinary, bringing together theories from Psychology, Neuroscience, Biology, and others. The field of cognitive science operates at a number of levels. On the one hand, some aspects of cognitive science look at the behaviour of neurons, whereas other aspects look to explain human behaviour at a more abstract level, seeking to explain human decision making or reasoning. A key area of research in cognitive science is how to formalise human behaviours around judgements of similarity, categorization, and decision making. In the field of Physics, Quantum Mechanics has fundamentally altered our understanding of the way in which particles behave. Quantum mechanics has a number of unintuitive phenomena, some of which can be used model unusual aspects of human behaviour. The application of quantum theory to model human behaviour is wide-ranging. In this chapter we will look at three main areas in which it has been applied. One key area is in how similarity judgements can be modelled. There are a number of phenomena around similarity judgements that are not well modelled using a view of concepts that does not take the state of the observer into account. These include the asymmetry of similarity judgements and the fact that similarity can change depending on other exemplars that are present. Another key area is in judgement and decision-making. Again, puzzling phenomena have been observed regarding the fact that judgements do not follow classical probability theory, known as the conjunction fallacy, can depend crucially on the order in which questions are presented or on the amount of knowledge about the world. Finally, approaches to modelling cognitive phenomena at the neural level will also be discussed within a quantum-theoretic framework.
Organizational decision making is often explored with theories from the heuristics and biases research program, which have demonstrated great value as descriptions of how people in organizations make decisions. Nevertheless, rational analysis and classical probability theory are still seen by many as the best accounts of how decisions should be made and classical probability theory is the preferred framework for cognitive modelling for many researchers. The focus of this work is quantum probability theory, an alternative probabilistic framework. Results in decision making, which appear paradoxical from a perspective of classical probability theory, may make perfect sense if one adopts quantum probability theory. We review some cognitive models of decision making based on quantum probability theory. Each of these models is based on a challenge to prescription from classical probability theory. The transition from labeling a particular behavior as irrational, by classical probability standards, to (potentially) rational (or, at any rate, not fallacious), raises interesting possibilities, including that of characterizing certain heuristics in formal, probabilistic terms.
Classical models of decision-making under uncertainty, grounded in expected utility theory and Bayesian probability, often fail to account for systematic cognitive anomalies such as preference reversals, order effects, ambiguity aversion, and context-dependent choices. Quantum cognition models offer a novel theoretical framework that applies the mathematical principles of quantum probability—such as superposition, interference, and non-commutativity—to cognitive processes without assuming physical quantum mechanisms in the brain. By representing cognitive states as probabilistic superpositions that collapse through decision “measurements,” these models provide parsimonious explanations for violations of classical rationality. This paper examines how quantum cognition frameworks capture contextual influence, belief updating, and interference effects in uncertain decision environments. Drawing on findings from cognitive psychology and behavioral economics, the study highlights the relevance of quantum cognition for applications in economic choice modeling, legal decision analysis, and artificial intelligence. The analysis suggests that quantum probability offers a robust alternative to classical approaches when modeling human decision-making in complex, ambiguous, and context-sensitive situations.