The Computer Science of Human Decisions

Establishing a unified theory of cognition has been an important goal in psychology^1,2. A first step towards such a theory is to create a computational model that can predict human behaviour in a wide range of settings. Here we introduce Centaur, a computational model that can predict and simulate human behaviour in any experiment expressible in natural language. We derived Centaur by fine-tuning a state-of-the-art language model on a large-scale dataset called Psych-101. Psych-101 has an unprecedented scale, covering trial-by-trial data from more than 60,000 participants performing in excess of 10,000,000 choices in 160 experiments. Centaur not only captures the behaviour of held-out participants better than existing cognitive models, but it also generalizes to previously unseen cover stories, structural task modifications and entirely new domains. Furthermore, the model’s internal representations become more aligned with human neural activity after fine-tuning. Taken together, our results demonstrate that it is possible to discover computational models that capture human behaviour across a wide range of domains. We believe that such models provide tremendous potential for guiding the development of cognitive theories, and we present a case study to demonstrate this.

Strategic decision-making is a crucial component of human interaction. Here we conduct a large-scale study of strategic decision-making in the context of initial play in two-player matrix games, analysing over 90,000 human decisions across more than 2,400 procedurally generated games that span a much wider space than previous datasets. We show that a deep neural network trained on this dataset predicts human choices with greater accuracy than leading theories of strategic behaviour, revealing systematic variation unexplained by existing models. By modifying this network, we develop an interpretable behavioural model that uncovers key insights: individuals’ abilities to respond optimally and reason about others’ actions are highly context dependent, influenced by the complexity of the game matrices. Our findings illustrate the potential of machine learning as a tool for generating new theoretical insights into complex human behaviours.

For much of the global population, climate change appears as a slow, gradual shift in daily weather. This leads many to perceive its impacts as minor and results in apathy (the ‘boiling frog’ effect). How can we convey the urgency of the crisis when its impacts appear so subtle? Here, through a series of large-scale cognitive experiments (N = 799), we find that presenting people with binary climate data (for example, lake freeze history) significantly increases the perceived impact of climate change (Cohen’s d = 0.40, 95% confidence interval 0.26–0.54) compared with continuous data (for example, mean temperature). Computational modelling and follow-up experiments (N = 398) suggest that binary data enhance perceived impact by creating an ‘illusion’ of sudden shifts. Crucially, our approach does not involve selective data presentation but rather compares different datasets that reflect equivalent trends in climate change over time. These findings, robustly replicated across multiple experiments, provide a cognitive basis for the ‘boiling frog’ effect and offer a psychologically grounded approach for policymakers and educators to improve climate change communication while maintaining scientific accuracy.