AI That Can Think, Reason and Discover: Markus J. Buehler

Conference Video|Duration: 25:15
April 1, 2025
Please login to view this video.
  • Video details

    AI That Can Think, Reason and Discover

    Markus J. Buehler
    Jerry McAfee Professor of Engineering, MIT Department of Civil and Environmental Engineering and MIT Department of Mechanical Engineering

    AI is evolving beyond pattern recognition into a tool for reasoning, discovery, and scientific insight. This talk explores how new AI architectures, including Reinforcement Learning (RL) and Graph Isomorphism Networks (GIN), enabling us to build powerful expressive AI models that move beyond memorization and into structural reasoning. By blending physics-driven models with generative AI, integrating biologically-inspired neural structures, and leveraging multi-agent systems that mirror collective intelligence in nature, we unlock new frontiers in scientific discovery. Case studies will highlight breakthroughs in materials science, demonstrating AI-driven advances with real-world applications in medicine, food, and agriculture. These developments showcase AI's potential not just as a tool for analysis but as an engine for reasoning, adaptation, and discovery, fundamentally reshaping our understanding of complex systems.

Locked Interactive transcript
Please login to view this video.
  • Video details

    AI That Can Think, Reason and Discover

    Markus J. Buehler
    Jerry McAfee Professor of Engineering, MIT Department of Civil and Environmental Engineering and MIT Department of Mechanical Engineering

    AI is evolving beyond pattern recognition into a tool for reasoning, discovery, and scientific insight. This talk explores how new AI architectures, including Reinforcement Learning (RL) and Graph Isomorphism Networks (GIN), enabling us to build powerful expressive AI models that move beyond memorization and into structural reasoning. By blending physics-driven models with generative AI, integrating biologically-inspired neural structures, and leveraging multi-agent systems that mirror collective intelligence in nature, we unlock new frontiers in scientific discovery. Case studies will highlight breakthroughs in materials science, demonstrating AI-driven advances with real-world applications in medicine, food, and agriculture. These developments showcase AI's potential not just as a tool for analysis but as an engine for reasoning, adaptation, and discovery, fundamentally reshaping our understanding of complex systems.

Locked Interactive transcript