
Sign up to save your podcasts
Or


Do models still matter when AI can already give us smart answers?In this episode of Supermodeling and AI, I explore why explicit models still matter—and how AI may actually make them more useful than ever.We return to RelayWorks, our fictional company facing a difficult AI product decision, and ask a capable AI what it should do. The answer sounds good. But does that actually resolve the decision?Max Henrion joins me to tell the amazing story of the award-winning California Rigs-to-Reefs project, where an explicit model helped break through years of disagreement among stakeholders.Then I share my own experience building the RelayWorks decision model with Assista—and why working with an AI modeler felt surprisingly different from the way I’ve built models for decades.If you’d like to try Assista yourself, it is included with the Free Edition of Analytica. Analytica currently requires Windows:https://analytica.com/#try-for-freeIf you enjoyed this topic, you will also like this presentation: From AI Advice to AI-Assisted Decision Models, https://youtu.be/Srx9J620wjY.Supermodeling and AI explores decision intelligence, modeling, and artificial intelligence—and what becomes possible when we bring them together.
What exactly is a model—and what do a balsa-wood airplane, an electrical circuit, a mental shortcut, a decision model, and a modern AI system have in common?In this first episode of Supermodeling and AI, we start with a simple idea: a model is a purposeful representation that leaves out complexity so we can do something useful with it.Along the way, I draw on a few personal stories—from building model airplanes as a kid, to studying electrical engineering, to developing an early neural-network system for English–Spanish translation in 1991—and connect them to the models we use every day, often without realizing it.We also introduce RelayWorks, a fictional software company deciding whether to launch, pilot, or delay a new AI feature called TechAssist. That case gives us a concrete way to ask four questions that apply to almost any model:• What is it for?• What comes out of it?• Where is its boundary?• What has it deliberately left out?We’ll also touch on framing—the process of deciding exactly what problem we are trying to solve—and on an important distinction that often gets blurred today: an AI model is not necessarily the same thing as an entire AI system.The goal of this episode is not to catalog every kind of model. It is to develop a way of thinking about models that will serve us throughout the series.And it leads directly to the next question:If generative AI can reason about a problem for us, do explicit models still matter?Supermodeling and AI explores decision intelligence, modeling, and artificial intelligence—and how they can work together to help us make better decisions.#DecisionIntelligence #AI #Modeling #DecisionMaking #ArtificialIntelligence
You’ve heard what AI knows. You’ve heard what it can do. Supermodeling and AI is about something else: what happens when AI helps you think.AI can generate answers, analyses, and recommendations with remarkable speed. But making a good decision still requires us to ask the right questions, understand uncertainty and tradeoffs, make our assumptions visible, and recognize what really matters.That’s why modeling isn’t becoming less relevant in the age of AI. It’s becoming even more important.Supermodeling and AI brings together decision intelligence, modeling, and artificial intelligence—using each to strengthen the others and help us reason through consequential decisions.I’m Lonnie Chrisman. My career has spanned both modeling and AI, from earning a Ph.D. in machine learning and artificial intelligence at Carnegie Mellon to decades spent developing modeling tools and applying them to difficult decisions.In this weekly series, you’ll learn practical ways to structure messy problems, reason more clearly under uncertainty, and use AI without simply handing your judgment over to it.Early episodes will explore:• What Is a Model?• Do Models Still Matter in the Age of AI?• What Is a Probability?• The Hidden AccompliceEpisodes generally run between fifteen and thirty minutes. Watch on YouTube for the diagrams and demonstrations, or listen through the podcast whenever that’s more convenient.The series launches November 3, 2026. Subscribe or follow to become a better modeler, a better user of AI, and a better decision maker.
From the publisher's feed