DSI Partners with SkAI and NITMB to Host Fifth Annual AI+Science Summer School

The 2026 AI+Science Summer School program introduced a new generation of interdisciplinary trainees to the new methodologies of AI+Science. This program has served to build communities across institutions and the physical sciences, encouraging collaborations and exchanges that spur new research directions focused on innovative AI-enabled scientific discovery.

The program was organized by the Eric and Wendy Schmidt AI in Science Fellowship program at the University of Chicago and the University of Chicago Data Science Institute. The AI + Science Summer School was co-hosted by the National Institute for Theory and Mathematics in Biology (NITMB) and SkAI Institute at the John Hancock Center in downtown Chicago.

This year’s program drew an international cohort of 76 physical and biological researchers spanning mathematical biology, molecular engineering, neuroscience, cosmology, medical imaging, biophysics, and more. Participants came from more than 30 universities or research institutions across North America, Europe, and Asia.

Cohort 4 Eric and Wendy Schmidt AI in Science Fellow and Summer School Organizing Committee member, Konstantin Gerbig said “It’s been great being able to be involved in organizing the summer school since conference organizing is an important component of almost any academic position, including those I am intending to pursue.” Konstantin added that “it has also been helpful in connecting with junior scientists from different disciplines, it informs my research and their’s, creating a mutual knowledge exchange.”

The week opened with an Organizing Committee welcoming address, then moved through five days of combined lectures and interactive tutorials led by national and international researchers, encouraging participants’ hands-on engagement with material. After the majority of session days, attendees had time to engage across disciplines and build a global AI+Science community at poster sessions, receptions, and dinners.

The Week’s Foundations and Materials

Features speakers from each day (with links to lectures) can be found below:

Monday, June 22: 

• Greg Shakhnarovich (TTIC / UChicago): “Representation Learning and Foundation Models for Images and Videos (Part I) (Part II).”
• Niall Mangan (Northwestern): “Sparse Model Selection for Dynamical Systems: Finding the Limits of Uncertainty (Part I) (Part II).”

Tuesday, June 23: 

• Earl Bellinger (Yale University): “Clustering Clusters: Searching for Star Clusters with the Gaia Spacecraft (Part I) (Part II).”
• Rose Cersonsky (University of Wisconsin): “Interpretability in AI/ML for Chemical Sciences (Part I) (Part II).”

Wednesday, June 24:

• Jonathan Weare (NYU): “Learning to Predict Extreme Events (Part I) (Part II).”
• Philippe Rigollet (MIT): “Optimal Transport and Transformer Flows with Applications to Single-Cell RNAseq Data (Part I) (Part II).”

Thursday, June 25:

• Emma Alexander (Northwestern): “AI for Imaging (Part I) (Part II).”
• Ramon Nogueira (UChicago): “Neuroscience and AI (Part I) (Part II).”

Friday, June 26:

• Bingqing Cheng (UC Berkeley): “Machine Learning Interatomic Potentials 101 (Part I) (Part II).”
• Francisco Villaescusa-Navarro (Flatiron Institute): “AI Agents for Science (Part I) (Part II).”

Across five days and a dozen speakers, a throughline emerged: AI is no longer a peripheral tool for science, but is becoming part of how discovery itself happens, whether that means finding star clusters in spacecraft data, predicting extreme events (i.e., climate, molecular, atmospheric changes), designing better materials, or understanding the brain.

Poster Session Connections

Across two poster sessions, fellows and students presented original research spanning the full breadth of the AI+Science community: from physics-informed neural networks for biomedical digital twin simulation, to optimal transport methods for inflammatory response prediction, to AI-boosted rare-event sampling for extreme weather forecasting.

Master’s student Rahma Aroua, said that “Presenting my AI and Neuroscience poster was one of the highlights of the summer school. I received valuable feedback from researchers across many scientific disciplines and had the opportunity to meet enthusiastic scientists whose diverse perspectives enriched my research and inspired new ideas.”

The range of topics on display, from cellular aging and protein-ligand binding to galaxy surveys and neuronal connectivity, underscored the program’s central premise: that AI-enabled discovery is taking shape simultaneously across nearly every scientific discipline, and that bringing those researchers into the same room accelerates all of it.

Eric and Wendy Schmidt AI in Science Fellowship Faculty Co-Director and Professor of Chemistry, Aaron Dinner, said “Our annual Summer School occupies a unique place in AI+Science in spanning all the way from introductory concepts to recent advancements,” adding that “the barriers to asking questions are low, facilitating learning and interactions between participants with different disciplinary backgrounds, making this program a productive space of growth and collaboration.”

As we look ahead to the sixth annual AI+Science Summer School, planning is already underway for Summer School 2027, which is slated to be held at the University of Chicago’s John W. Boyer Center in Paris, France.

Written by Casey Keel, DSI Scientific Writer.