Supporting STEM Education in Northern New Mexico
How Agentic AI is Supercharging Science at LANL Talk by Dr. Nathan DeBardeleben
NEW: Free child care provided at the STEAM Lab - spaces for up to 15 children, register separately via Google Form.
In this talk, Dr. DeBardeleben will discuss how AI agents are changing science by aiding in automating tasks and empowering the scientist in discovery. Nathan will discuss agentic tools used in industry and programming and then will present URSA, a LANL open source agent framework for science. The talk will focus on use cases for this technology and will include video examples of AI agents working on LANL tasks.
Artificial intelligence is rapidly moving beyond passive assistance toward systems that can reason, plan, use tools, write code, run analyses, and help scientists move from question to result more quickly. In this talk, Dr. Nathan DeBardeleben will discuss how agentic AI is beginning to transform scientific workflows at Los Alamos National Laboratory by automating routine tasks, amplifying expert judgment, and enabling new modes of discovery. The talk will begin with a brief overview of agentic AI tools in industry and software development, then shift to scientific applications at LANL. Nathan will introduce URSA, a LANL open-source agent framework for science, and highlight how agent-based systems can support tasks such as information retrieval, coding, analysis, simulation workflows, and experiment planning. The presentation will emphasize real use cases and will include video examples of AI agents working on LANL-relevant tasks.
$10 admission; free for BSMA members (you can become a member on our Join Us page).
Wednesday, July 8, 6-8pm at the Bradbury Science Museum. Free childcare available!
About the Speaker
Dr. Nathan DeBardeleben is a Senior HPC Architect at Los Alamos National Laboratory, where he has worked since 2004 after earning his Ph.D. in Computer Engineering from Clemson University. He spent more than a decade studying supercomputer reliability and the effects of cosmic radiation before shifting into applied machine learning for LANL’s Enabling Manufacturing effort. Since then, Nathan and his team have applied machine learning to high-impact production challenges in support of national security missions. Nathan currently serves as co-PI of AI for Mission, or ArtIMis, a Signature Institutional Commitment LDRD effort. ArtIMis represents a $20 million per year investment in frontier artificial intelligence technologies designed to accelerate DOE mission science.
The Bradbury Science Museum Association supports and inspires learners of all ages in Northern New Mexico and beyond through STEM Education.
