Event Details:
Save the Date: May 7, 2025
The 2025 Symposium of the Center for Mind, Brain, Computation and Technology (MBCT) focuses on a striking disparity between natural and artificial intelligence.
While AI systems are approaching or exceeding ordinary humans in many tasks, humans—and sometimes non-human animals—are far more efficient learners. Today’s AI systems require roughly 100,000 times more experience than a human learner could receive in a lifetime. Our symposium asks: What approaches should we take to address this gap?
Organized by Jay McClelland, Director of MBCT, the symposium brings together scientists from the fields of systems, cognitive, and computational neuroscience and from artificial intelligence research to explore these questions. Three of the speakers have investigated mechanisms of learning in biological brains, including a recently discovered form of plasticity called Behavioral Time Scale Synaptic Plasticity, as well as the role of experience replay in enhancing the effectiveness and efficiency of learning. The symposium will also showcase research in artificial intelligence that captures a powerful capability shared by machines and animals alike: the ability to learn to learn – becoming more efficient learners as they accumulate experience.
The symposium will begin with a brief overview by Jay McClelland, followed by talks from:
- Bruce McNaughton (University of California, Irvine & University of Lethbridge)
- Aaron Millstein (Rutgers University)
- Anna Schapiro (University of Pennsylvania)
- Feryal Behbahani (Google DeepMind)
The symposium will conclude with a moderated discussion featuring faculty affiliates of the Wu Tsai Neurosciences Institute and questions from the audience. A closing reception from 5:00–6:00 PM will provide opportunities for continued conversations.
Sign up to receive future updates about the symposium and additional events