{"type":"rich","version":"1.0","provider_name":"Transistor","provider_url":"https://transistor.fm","author_name":"Runpoint: AI Business Transformation Podcast","title":"Local Models, Open Weights, and AI Sovereignty: A Practitioner Roundtable","html":"<iframe width=\"100%\" height=\"180\" frameborder=\"no\" scrolling=\"no\" seamless src=\"https://share.transistor.fm/e/49a5a770\"></iframe>","width":"100%","height":180,"duration":2383,"description":"Sam Gaddis sits down with the largest Runpoint Podcast panel yet: Matthew Hall and Ryan Mish from the Runpoint team, Thomas McNally of Zaelab, and returning guest Thanh Pham.\nThe conversation starts with the Theo video that stirred up the \"local models are overrated\" argument and goes deep from there. Thomas walks through what it actually looks like to deploy open weight models across a firm, including the cost math, the hardware, and why non-technical employees are lining up to join the program. Thanh brings the hobbyist-turned-practitioner view from his home lab and where local genuinely holds up. Along the way: the four-way split between frontier, local-on-prem, local-in-cloud, and rented open weight inference, plus a straight look at the two fears CEOs raise most often.\nIf you run a mid-market company and you're trying to figure out where these models fit, this one is for you.\nGuests:\nThomas McNally, technology lead at Zaelab\nThanh Pham, Managing Director of Asian Efficiency\nMatthew Hall and Ryan Mish, Runpoint\nTimestamps:\n0:00 Intro and the panel\n0:53 The Theo video and the local vs open weight debate\n1:22 Deploying open weight models across a firm: the cost story\n4:41 Which models and hardware Zaelab landed on (Qwen, RTX 6000, VRAM)\n6:41 Why employees want in on an \"inferior\" model program\n8:00 The interface problem and Anthropic's third-party inference feature\n9:28 Pay-per-use to all-you-can-eat, and protecting your IP\n9:56 The subscription gravy train is ending\n10:21 AI sovereignty and why the conversation shifted\n12:11 Thanh's home lab and the hybrid local setup\n13:41 What \"most people\" can actually run on their hardware\n16:03 The routing trifecta: frontier, cloud infrastructure, and on-device\n18:32 What each panelist actually uses day to day\n19:30 Model agnosticism and avoiding vendor lock-in\n21:40 How to build an intelligent routing layer\n24:14 Evaluating routing solutions without a lab\n24:54 Myth busting: Chinese models and training on your data\n27:18...","thumbnail_url":"https://img.transistorcdn.com/J8t_zr9ARjI2f2cdFABJsSzRDnvqZ_RmHIhyQWPPIXs/rs:fill:0:0:1/w:400/h:400/q:60/mb:500000/aHR0cHM6Ly9pbWct/dXBsb2FkLXByb2R1/Y3Rpb24udHJhbnNp/c3Rvci5mbS9jZjVl/ZmRjZDdmZjBiNDY3/ZjNjMWQxY2ZkMDY3/Y2FiNS5qcGc.webp","thumbnail_width":300,"thumbnail_height":300}