AI Science Center Seminar: The AlphaFold Moment That Hasn't Happened (Yet) - AI Meets the LHC

27.04.2026 11:00 – 12:00

Particle physics sits at an inflection point. After a decade of applying machine learning to individual tasks: jet classification, fast simulation, anomaly detection, the field has accumulated impressive but incremental gains. No "AlphaFold moment" has arrived. Meanwhile, AI itself has undergone a phase transition: frontier models now function as genuine research collaborators, and agentic AI systems can autonomously plan, execute, and iterate on complex scientific workflows. This talk argues that the convergence of three developments — physics-native foundation models, simulation-based inference, and agentic AI — creates a unique window of opportunity to fundamentally redesign how we do experimental physics. I will review the current landscape of ML in HEP through the lens of the HEPML Living Review, diagnose the organizational pathologies that have prevented transformative progress, and present a concrete vision for an AI-native analysis pipeline that takes us from raw detector data to publication-ready physics results. The talk culminates with a provocation: a specific "moonshot" research program — scalable, autonomous discovery of new narrow resonances across the full LHC dataset — and an honest assessment of what it would take. I will close with an open discussion on how the AISC community, with its unique interdisciplinary composition, can contribute to making this vision real.

Lieu

Bâtiment: Battelle

- Amphitheater D
- ground floor hall of Battelle A for coffee and snacks (starting at 10:30)

entrée libre