The Physical AI Stack: Where Founders Should Build

Physical AI is not won by hype, it is won by the companies that understand where the real value sits. In this this episode of robotbuildersclub.org, Ali Afzal hosts Bibhrajit Halder, Founder of SafeAI, and breaks down why the biggest opportunities in robotics are not just in the robot itself, but across manufacturing, raw materials, models, infrastructure, workflow, distribution, and service. Ali Afzal and Bibhrajit return for a wide-ranging, highly practical conversation about the future of robotics, autonomy, and physical AI. If you are a founder, engineer, investor, or operator trying to figure out where to place your time, capital, or next startup bet, this episode gives you a grounded map of the space. Bibhrajit explains the physical AI stack from the ground up: hardware, compute, software, infrastructure, and applications and why founders should think beyond the obvious layer. He unpacks why manufacturing and mining may be hidden value centers under hardware, why most startups cannot afford to build foundation models from scratch, and why open source plus proprietary vertical data is likely to dominate many real-world use cases. Takeaways Physical AI presents significant opportunities in the manufacturing and construction industries. The role of hardware, manufacturing, and open source models is crucial in the development of physical AI applications. The future of robotics and physical AI will involve a working model that emphasizes partnerships between technology providers and end users. Techno-optimism and anxiety about job displacement coexist, but the macro view shows that job opportunities and progress continue to grow. Investment in physical AI is influenced by market trends, technical moats, and the founders leading the companies. Exit optionality is important for robotics and physical AI companies, and pressure plays a role in considering an exit. M&A success is closely tied to profitability A technological breakthrough is needed to advance physical AI and robotics Chapters 00:00 Introduction to Physical AI and Robotics 03:23 The Value of Physical AI in Manufacturing 06:10 Hardware and Manufacturing in Physical AI 08:01 Building and Training Models in Physical AI 09:24 Open Source vs. Closed Source Models 13:36 Foundation Model in Physical AI 19:04 The Opportunity in Construction Use Case 25:18 The Role of NVIDIA and Other Players in Physical AI 32:39 The Real Opportunity in Construction 36:38 The Future of Robotics and Physical AI 42:21 The Impact of Robotics and AI on Society 54:15 Funding and Investment in Physical AI 01:02:34 Exit Strategies for Robotics and Physical AI Companies 01:08:29 The Challenge of M&A 01:17:13 The Future of Physical AI and Robotics