103: Using AI in Manufacturing: Generative vs. Predictive and Autonomous AI with Bryan DeBois

Send us Fan MailMost manufacturers are chasing the wrong AI problem. In this episode Chris talks with Bryan DeBois, Director of Industrial AI at RoviSys, about why industrial AI for manufacturing requires a different approach than generative AI.Bryan explains the limits of generative AI on the plant floor, why deterministic systems matter in high risk environments, and how analytical AI, predictive AI, computer vision, and autonomous AI are already being used to improve quality, safety, throughput, and asset performance. Leaders should listen to understand how industrial AI can protect expertise, strengthen operations, and create practical advantage beyond the ChatGPT conversation.Chapters:00:00 Introduction01:16 Why Factory Floor AI Is Different From Knowledge Work AI03:58 The Four Types of AI Used in Manufacturing Today05:50 Why Generative AI Fails in High Risk Operational Environments10:56 Manufacturing Risks Also Apply to Construction and Life Sciences11:36 The Workforce Crisis Driving Industrial AI Adoption15:26 Why Manufacturing Careers May Be Safer Than White Collar Jobs20:42 Why Humanoid Robots Are Not the Future of Manufacturing24:53 Capturing Tribal Knowledge Before Experts Retire40:00 Who Should Own AI Inside Manufacturing Organizations43:24 Meta’s Cicero Project and the Future of Hybrid AI Systems47:08 Deterministic AI vs Probabilistic AI in Critical Industries49:27 Where to Follow Brian De Bois and Learn More About Industrial AIResources:🔎 Find Out More About Bryan DeBoisBryan DeBois on LinkedIn:https://www.linkedin.com/in/bryan-deboisRoviSys Industrial AI: https://www.rovisys.com/aiRoviSys:https://www.rovisys.com🛠 AI Tools and Resources Mentioned:ChatGPT: https://chatgpt.com/Claude: https://claude.com/Grok: https://grok.com/Meta AI CICERO: https://ai.meta.com/research/ciceroGoogle DeepMind AlphaGo: https://deepmind.google/research/breakthroughs/alphagoMicrosoft HoloLens: https://www.microsoft.com/hololensObsidian: https://obsidian.mdSAP:https://www.sap.com