What AI is teaching us about the human mind
By Wendy Sutton
Large language models were built to predict the next word in a sequence, but their surprising similarities with the human mind are taking cognitive science in bold new directions.
Researchers at the University of Michigan are using those parallels to better understand complex human abilities such as language use and reasoning, and, ultimately, to shed new light on creativity, morality and mental health. Richard Lewis and Chandra Sripada study the computational basis of the mind and brain. Lewis, Arthur F. Thurnau Professor and John R. Anderson Collegiate Professor of Psychology, Linguistics and Cognitive Science, examines language processing, decision-making and skilled performance. Sripada, Theophile Raphael Research Professor of Clinical Neurosciences and professor of psychiatry and philosophy, focuses on executive functions, neurodevelopment and decision-making.
Over the past five years, the advent of generative AI has provided the first working examples of how certain complex mental functions may operate in areas of the mind and brain that are otherwise difficult to study. These models are giving researchers new testable hypotheses.
“We’re discovering a lot more convergence between how the human mind and new AI systems work than you might expect. In some ways, that’s surprising because AI hasn’t lived a life like a human. On the other hand, it’s not surprising because AI and cognitive science have a long, rich history of working together. Many of the core ideas these systems were trained on came out of cognitive science, including some of the important ideas behind neural networks.”
“We’re discovering a lot more convergence between how the human mind and new AI systems work than you might expect,” Lewis said. “In some ways, that’s surprising because AI hasn’t lived a life like a human. On the other hand, it’s not surprising because AI and cognitive science have a long, rich history of working together. Many of the core ideas these systems were trained on came out of cognitive science, including some of the important ideas behind neural networks.”
Despite advances in brain imaging and neural recording, much of what is known about language processing still comes from eye tracking. As readers move through a sentence, they must retain earlier information, such as the subject of a verb or the referent of a pronoun. In reading-time studies, participants read text while researchers record how long their eyes linger on each word. The data show where a reader is struggling and what memory processes are involved.
To better understand this working memory, or short-term memory in language processing, Lewis leads transformer reading-time studies. The transformer architecture, the design behind tools like ChatGPT, processes sentences one word at a time by computing relations between each word and what came before.
Because that resembles how humans are believed to handle language in working memory, Lewis examines the transformer’s internal memory-retrieval patterns, rather than just its output, as it processes sentences. Those patterns predict how long a participant’s eyes will spend on particular words, even though the models were never trained for that purpose.
Sripada’s research examines fast and slow thinking, a dual-process concept in which automatic processing routes allow for fast thinking, while deeper processing supports slow thinking.
Cognitive scientists study fast and slow thinking using conflict tasks, such as the Stroop task. Participants read color words printed in potentially conflicting ink colors. For example, the word “red” might be printed in blue ink and they must name the ink color. Reading the word is the more practiced response, so the brain’s fast processing tends to produce “red.” Naming the ink color requires slower processing.
Surprisingly, LLMs have developed a similar duality, even though they were just trained to predict the next word. They possess automatic processing as well as in-context processing, or the ability to draw on deeper resources and detect hidden patterns in a prompt.
“That’s shocking, because I would have thought dual process structure was evolutionarily shaped and is essentially hard-wired,” Sripada said. “This changes our understanding of what this duality is and how it emerges. We’re still studying the same phenomenon we always have, but with a new perspective on what it is and how it emerged.”
“That’s shocking, because I would have thought dual process structure was evolutionarily shaped and is essentially hard-wired. This changes our understanding of what this duality is and how it emerges. We’re still studying the same phenomenon we always have, but with a new perspective on what it is and how it emerged.”
The findings may carry important implications for developmental psychology and the study of mental disorders. Sripada hopes the research will uncover insights into conditions such as schizophrenia, depression, ADHD and substance abuse, where the balance between fast and slow thinking may be disrupted. Sripada plans to use fMRI to scan participants as they complete conflict tasks and compare the resulting brain activation patterns with the model’s patterns to see where they align.
This work also raises questions about human creativity, agency, morality and perhaps even consciousness. Researchers have long lacked clear mechanistic models for how people generate creative products, such as poems or stories. LLMs are now capable of producing works of substantial novelty, suggesting creativity can emerge from prediction-based learning itself. A similar strategy may shed light on goal-directed control, self-awareness, metacognition and other sophisticated human abilities.
Sripada directs the Weinberg Institute for Cognitive Science, which Lewis formerly directed. As part of a faculty expansion program sponsored by the provost’s office, U-M awarded a four-person faculty cluster in AI Cognitive Science involving Psychology, Linguistics, Philosophy and Psychiatry. The faculty will have partial appointments in the institute. The cluster will expand AI cognitive science research at U-M, helping position the university as a leader in the field.
“The technology of AI is what has been making the news,” Lewis said. “But it’s not just a tech story. People don’t realize that there’s also a revolution happening in cognitive science, and this is by far the most exciting time to be in this field.”