David A. Lagnado

Evangelia Kyrimi, Scott McLachlan, Jared M Wohlgemut et al.

Proposals of artificial intelligence (AI) solutions based on increasingly complex and accurate predictive models are becoming ubiquitous across many disciplines. As the complexity of these models grows, transparency and users' understanding often diminish. This suggests that accurate prediction alone is insufficient for making an AI-based solution truly useful. In the development of healthcare systems, this introduces new issues related to accountability and safety. Understanding how and why an AI system makes a recommendation may require complex explanations of its inner workings and reasoning processes. Although research on explainable AI (XAI) has significantly increased in recent years and there is high demand for XAI in medicine, defining what constitutes a good explanation remains ad hoc, and providing adequate explanations continues to be challenging. To fully realize the potential of AI, it is critical to address two fundamental questions about explanations for safety-critical AI applications, such as health-AI: (1) What is an explanation in health-AI? and (2) What are the attributes of a good explanation in health-AI? In this study, we examined published literature and gathered expert opinions through a two-round Delphi study. The research outputs include (1) a definition of what constitutes an explanation in health-AI and (2) a comprehensive list of attributes that characterize a good explanation in health-AI.

Maria Victoria Carro, David Lagnado

AI-related incidents are becoming increasingly frequent and severe, ranging from safety failures to misuse by malicious actors. In such complex situations, identifying which elements caused an adverse outcome, the problem of cause selection, is a critical first step for establishing liability. This paper investigates folk perceptions of causal responsibility in causal chain structures when AI systems are involved in harmful outcomes. We conduct human experiments to examine judgments of causality, blame, foreseeability, and counterfactual reasoning. Our findings show that: (1) When AI agency was moderate (human sets the goal, AI determines the means) or high (AI sets the goal and the means), participants attributed greater causal responsibility to the AI. However, under low AI agency (where a human sets both a goal and means) participants assigned greater causal responsibility to the human despite their temporal distance from the outcome and despite both agents intended it, suggesting an effect of autonomy; (2) When we reversed roles between human and AI, participants consistently judged the human as more causal, even when both agents perform the same action; (3) The developer, despite being distant in the chain, was judged highly causal, reducing causal attributions to the human user but not to the AI; (4) Decomposing the AI into a large language model and an agentic component showed that the agentic part was judged as more causal in the chain. Overall, our research provides evidence on how people perceive the causal contribution of AI in both misuse and misalignment scenarios, and how these judgments interact with the roles of users and developers, key actors in assigning responsibility. These findings can inform the design of liability frameworks for AI-caused harms and shed light on how intuitive judgments shape social and policy debates surrounding real-world AI-related incidents.

Magnus F. Gjerde, Vanessa Cheung, David Lagnado

With the rapid progress of Large Language Models (LLMs), it becomes increasingly important to understand their abilities and limitations. In two experiments, we investigate the causal and compositional reasoning abilities of LLMs and humans in the domain of object affordances, an area traditionally linked to embodied cognition. The tasks, designed from scratch to avoid data contamination, require decision-makers to select unconventional objects to replace a typical tool for a particular purpose, such as using a table tennis racket to dig a hole. In Experiment 1, we evaluated GPT-3.5 and GPT-4o, finding that GPT-4o, when given chain-of-thought prompting, performed on par with human participants, while GPT-3.5 lagged significantly. In Experiment 2, we introduced two new conditions, Distractor (more object choices, increasing difficulty) and Image (object options presented visually), and evaluated Claude 3 Sonnet and Claude 3.5 Sonnet in addition to the GPT models. The Distractor condition significantly impaired performance across humans and models, although GPT-4o and Claude 3.5 still performed well above chance. Surprisingly, the Image condition had little impact on humans or GPT-4o, but significantly lowered Claude 3.5's accuracy. Qualitative analysis showed that GPT-4o and Claude 3.5 have a stronger ability than their predecessors to identify and flexibly apply causally relevant object properties. The improvement from GPT-3.5 and Claude 3 to GPT-4o and Claude 3.5 suggests that models are increasingly capable of causal and compositional reasoning in some domains, although further mechanistic research is necessary to understand how LLMs reason.

Neil R. Bramley, Peter Dayan, Thomas L. Griffiths et al.

Higher-level cognition depends on the ability to learn models of the world. We can characterize this at the computational level as a structure-learning problem with the goal of best identifying the prevailing causal relationships among a set of relata. However, the computational cost of performing exact Bayesian inference over causal models grows rapidly as the number of relata increases. This implies that the cognitive processes underlying causal learning must be substantially approximate. A powerful class of approximations that focuses on the sequential absorption of successive inputs is captured by the Neurath's ship metaphor in philosophy of science, where theory change is cast as a stochastic and gradual process shaped as much by people's limited willingness to abandon their current theory when considering alternatives as by the ground truth they hope to approach. Inspired by this metaphor and by algorithms for approximating Bayesian inference in machine learning, we propose an algorithmic-level model of causal structure learning under which learners represent only a single global hypothesis that they update locally as they gather evidence. We propose a related scheme for understanding how, under these limitations, learners choose informative interventions that manipulate the causal system to help elucidate its workings. We find support for our approach in the analysis of four experiments.