In this episode, Will Summerlin is joined by Lin Qiao who shares her journey from leading the PyTorch team at Meta to founding Fireworks, a company focused on specialized, customized AI models for enterprise companies and startups. They explore the challenges of training and inference costs in building AI infrastructure, the potential of multimodal models, and the strategic shift from 'one model fits all' to a diverse set of specialized models. They also cover how to think about efficiency in model deployment, the advantages of open-source vs. closed-source and why her philosophy is opposite to OpenAI's. 🔥 Apply to join over 400 founders and Execs in the Turpentine Network: https://hmplogxqz0y.typeform.com/to/JCkphVqj RECOMMENDED PODCAST: 🎙️ Second Opinion - A new podcast for health-tech insiders from Christina Farr of the Second Opinion newsletter. Join Christina Farr, Luba Greenwood, and Ash Zenooz every week as they challenge industry experts with tough questions about the best bets in health-tech. Apple Podcasts: https://podcasts.apple.com/us/podcast/id1759267211 Spotify: https://open.spotify.com/show/0A8NwQE976s32zdBbZw6bv - 📈 ECON 102 W/ NOAH SMITH - Keeping econ entertaining - Noahpinion author Noah Smith and Erik Torenberg break down business news and politics every week. Spotify: https://open.spotify.com/show/3Mos4VE3figVXleHDqfXOH Apple: https://podcasts.apple.com/id1753399812https://podcasts.apple.com/id1753399812 -- LINKS: https://fireworks.ai/ X/SOCIAL: @WillSummerlinAI (Will) @lqiao (Lin) TIMESTAMPS: (00:00) Preview and Intro (02:22) From Meta to Building Fireworks AI (04:29) Challenges and Innovations in AI at Scale (06:53) Navigating the Tidal Wave of Generative AI (08:27) The Evolution from Large to Specialized Models (16:30) Open Source vs. Closed Source (31:33) Optimizing AI from Training to Inference Costs (37:54) The Future of Customization (44:48) Customer Obsession (46:20) Wrap