"Reason and Inquiry is a major contribution to the philosophy of mind, the psychology of reasoning, and cognitive science, with implications for linguistics, epistemology, and decision theory. The erotetic theory looks set to be a key player in future debates on the nature of rationality."
Wykeham Professor of Logic
University of Oxford
"An insightful treatment of reason and rationality, explaining many puzzles and integrating many viewpoints."
Johnstone Professor of Psychology
"It is easy for researchers in Artificial Intelligence (AI) to get excited with our technical achievements and lose track of the big questions: what is intelligence, and how does it work? This thought provoking and wide-ranging book prompts us to look again at our field: to revisit the most basic questions surrounding our endeavour, and, perhaps most importantly, to consider new directions for the future."
Co-Director for AI
The Alan Turing Institute
Fulford Clarendon Professor
of Philosophy and Cognitive Science
University of Oxford
Senior Research Associate
Institute for Ethics in AI
University of Oxford
St. Catherine's College
Ph.D. (Philosophy and Neuroscience) Princeton University
B.A. Pomona College
I work on the human capacity for reasoning and decision-making, and how it relates to artificial agents and large language models like GPT. My recent book Reason and Inquiry presents a theory of this capacity and its two-faced nature: On the one hand, we are subject to systematic fallacies and framing effects, empirically documented in psychology and behavioural economics. On the other hand, we largely get things right and are capable of incredible feats of rationality. I argue that our minds naturally aim at resolving issues or answering questions as directly as possible, and if we are inquisitive enough in the process, we can get the kind of rationality required for science, philosophy, and classical economic agents as a special case. I am also interested moral judgment and in definitions of intelligence, both in humans and in artificial intelligence (AI). I regularly collaborate with computer scientists on these topics. I have also had many past collaborations with colleagues from psychology, linguistics, neuroscience, and environmental systems.
May 4, 2023 - Lunch talk at Imbue (Generally Intelligent) in San Francisco.
May 4, 2023 - Talk with Vincent Wang-Maścianica at Thomas Icard's Language Group at Stanford.
April 25, 2023 - First meeting of grad seminar with Will Davies, "Topics in Minds and Machines: Perception, Cognition, and ChatGPT." 11am Ryle Room.
March 30, 2023 - New Paper (with Vincent Wang-Maścianica): We found production of human-like fallacious judgments to increase from GPT-3 to GPT-4, even as it got much better at human-like correct judgments too. Perhaps no surprise, since GPT is trained on human text. #HumansInHumansOut https://arxiv.org/abs/2303.17276
January 17, 2023 - First meeting of "A Theory of Reason" graduate seminar with Sean Moss (Computer Science)
AI, Reasoning and Decision-Making
Madsen, J., Bailey, R., Carrella, E., Koralus, P. (2019). “Analytic versus computational cognitive models: Agent-based modeling as a tool in cognitive sciences.” Current Directions in Psychological Science, 28(3), pp. 299-305.
Koralus, P. and Mascarenhas, S. (2018). “Illusory inferences in a question-based theory of reasoning.” In: Horn, L. and Turner, K. (Eds.) An Atlas of Meaning (Current Research in the Semantics/Pragmatics Interface). Brill.
Mascarenhas, S. and Koralus, P. (2015). “Illusory inferences: disjunctions, indefinites, and the erotetic theory of reasoning.” In: Noelle, D. C., et al. (Eds.). Proc. 37th Meeting of the Cognitive Science Society, Cognitive Science Society.
Koralus, P. and Mascarenhas, S. (2013). “The erotetic theory of reasoning: Bridges between formal semantics and the psychology of propositional deductive inference.” Philosophical Perspectives, 27, pp. 312-365.
Moral judgment and Delusional thinking
(commentary by John McKay and symposium on http://imperfectcognitions.blogspot.co.uk)
Parkinson, C. Sinnott-Armstrong, W., Koralus, P., Mendelovici, A., McGeer, V. and Wheatley, T. (2011). “Is morality unified? Evidence that distinct neural systems underlie moral judgments of harm, dishonesty, and disgust.” Journal of Cognitive Neuroscience, Vol. 23(10), pp. 3162-80.
Attention and Perception
Koralus, P. (2016) “Can visual cognitive neuroscience learn anything from the philosophy of language? Ambiguity and the topology of neural network models of multistable perception.” Synthese, 193(5), pp. 1409-32.
(selected for symposium at www.philosophyofbrains.com, replies by Chris Mole, Felipe de Brigard, Sebastian Watzl, et al.)
I regularly offer the following graduate seminars:
A Theory of Reason: Philosophy, Psychology, and Algorithms
Topics in Minds and Machines: Perception, Cognition, and ChatGPT
I regularly offer undergraduate instruction on the following, among other topics:
Philosophy of Cognitive Science
Philosophy of AI
I have offered doctoral supervision in both philosophy and computer science in a variety of topics including language, consciousness, and moral judgment.