Agenda

Infrastructure and construction - spanning energy, data center, transport, water, cities, and supply chains - is the world’s largest industry, yet it remains one of the least cognitively equipped. Despite decades of projects, failures, delays, and geopolitical shocks, infrastructure decision-making largely operates without institutional memory. This talk introduces the concept of cognitive infrastructure: the layer of data, models, and human-AI judgment systems that enables societies and organizations to remember, learn, and reason under deep uncertainty. Drawing on work at the intersection of AI, reliability engineering, and infrastructure economics, the talk argues that the core challenge is not automation but critical thinking, judgment, and decision quality in complex, high-stakes environments. Using examples from global infrastructure planning, supply chains, and geopolitical risk, the talk demonstrates how modern AI when embedded as a reasoning and memory layer can help engineers, policymakers, and executives make better decisions in the presence of uncertainty, incomplete data, and long time horizons.