Event: Department Seminar

Location: Virtual

Contact Person: Juan Carlos Cordoba

Presentation Title: Completely Agent-Based Modeling (c-ABM): A Right Mathematics for Coupled Physical and Economic Systems?

Presenter: Leigh Tesfatsion (Virtual)

Date/Time of Presentation: Friday, Feb 28, 3:40pm – 5:00pm

Link for Presentation Slide-Set (pdf): https://www2.econ.iastate.edu/tesfatsi/CompletelyABM.LTesfatsion.AIWorkshopISU2025.pdf

Presentation Abstract:

The increasingly rapid progression from AI assistants to data-driven generative AI agents and beyond (e.g., embodied untethered agents with encapsulated data-driven generative AI capabilities) could be one of the greatest technological drivers for societal change since the advent of commercially available web browsers during the mid-1990s.  What role do economists envision for themselves in this turbulent process?

This talk will first present, in axiomatized form, a modeling approach -- called completely Agent-Based Modeling (c-ABM) – that permits the study of coupled physical and economic systems with AI-supported decision makers.  More precisely, it will be shown how c-ABM permits the modeling of decision-makers with encapsulated intentionality informed by data-driven generative AI.  It will next be shown how c-ABM supports high-fidelity science-with-practice studies of real-world systems taking a coupled physical and economic form for which physical devices as well as humans have autonomous AI capabilities, such as grid-supported electric power markets.  The talk will conclude by showing how c-ABM provides important support for Iterative Participatory Modeling (IPM), an open-ended collaborative learning process among stakeholders and researchers to bridge the gap between concept and practice.