The authors treat understanding as a relation linking a system and a target, and then show the many ways that relation can be specified. This matters for engineers, evaluators, and policymakers who must decide what kinds of tests, explanations, or safeguards are appropriate. Thinking carefully about the system, the target, and the relation helps avoid vague claims and points teams toward measurable, actionable questions.
For anyone curious about human potential and inclusion, the framework offers a way to connect technical choices to real-world effects. Clearer definitions of machine understanding can change who benefits from AI, how errors are caught, and which skills humans continue to develop alongside machines. Follow the link to see the options the authors map out and to consider how those options might shape the future of intelligent tools in everyday life.
What do artificial intelligence (AI) systems “understand”? This question arises not only in assessing a system’s intelligence but also in evaluation practices to ensure the safe and responsible deployment of AI. Drawing on scholarship from philosophy and cognitive science, and informed by current practices in AI, we develop a framework for asking more precise questions and making more precise claims about machine understanding. We conceptualize understanding as a relation between a system (S) and a target of understanding (T), and we discuss how to specify the relation, the system, and the target, offering a landscape of options in each case. Our goal is not to defend a particular account of understanding, but to provide conceptual tools for those working to assess or advance machine understanding.
