Research
Understanding the uniqueness of human intelligence remains one of the defining intellectual challenges of our time. Across history, thinkers have sought to formalize how human learners acquire concepts, structure the natural world, and pass on knowledge through culture and pedagogy. Today, this foundational inquiry is profoundly amplified, and indeed challenged, by the rapid breakthroughs of artificial intelligence. Recent milestones force urgent questions: How does a machine’s success in abstract domains compare to a human’s? If machines think differently, can human intelligence be coded into them to produce artificial human cognition? Addressing these questions is critical to our understanding of AI’s results, as exemplified by AI’s recent triumph in a Millennium Prize Problem, which left human mathematicians to wonder whether AI’s proof would be scrutable by us humans and thereby lead to greater human understanding of mathematics. Or consider the recent AI escapes from their “sandboxes,” which left us all wondering whether this was the work of ethical agents or merely task-driven bots. In the current crescendo, we should seek AI that we can understand and that better understands us.
Our research is uniquely positioned to address such questions and achieve such goals, while also shedding new light on the perennial goal to elucidate human cognition. We investigate the origins and development of two quintessential domains of human cognition: geometric cognition and social cognition. We deploy perceptual and conceptual experiments with infants and children, language-learning paradigms, computational modeling, and large-scale field interventions. We have also pioneered a new paradigm that adapts AI methodology to probe the foundations of human intelligence. The totality of our work explores how deep-seated geometric and social intuitions ground abstract human thought and how human developmental milestones can serve as rigorous blueprints for evaluating and advancing machine intelligence.
We have active collaborations with economists, mathematicians, neuroscientists, educators, and humanists, and we have research partnerships with The National Museum of Mathematics (MoMath), Lookit (the online infant and child lab), and the Abdul Latif Jameel Poverty Action Lab (J-PAL).
If you’re interested in joining our lab as a doctoral student, click here to find out more information about applying to the program in Cognition & Perception at NYU.
