
The AI Model Wars. Who wins - and What Should Founders Choose? From the former head of OpenAI Inference
Send us Fan Mail In this episode of Startup Spotlight, hosts Shirish Nadkarni and Gowri Shankar speak with Anuj Gosalia, former Head of Inference at OpenAI and a veteran engineering leader at Microsoft, Meta, and Google. Drawing on his experience at the forefront of GPUs, virtual reality, and AI, Anuj explores how large language models are reshaping entrepreneurship—and why solving real customer problems remains the foundation of a successful startup. Key topics covered include: Frontier vs. open-weight models : How the capability gap is narrowing and why benchmark scores don’t always reflect real-world performance. Choosing the right AI model : Why Anuj recommends using the best available models for coding, then selecting models for your product through evaluations tailored to customer needs. Product-market fit before cost optimization : Why founders should first build something customers value, then optimize inference costs as usage grows. Distillation and enterprise knowledge : How learning from AI interactions can help companies build institutional knowledge and improve their systems. Chinese models and customer trust : The distinct considerations around model origins, inference providers, hosting locations, and data governance. AI regulation and societal impact : Anuj’s perspective on accountability, geopolitical competition, inequality, and the challenges of rapid technological change. Coding tools, agents, and engineering culture : How integrated AI tools and customer usage create competitive advantages, and why faster coding and agent swarms still leave complex engineering problems to solve. Brought to you by TiE Seattle Hosts: Shirish Nadkarni and Gowri Shankar Producers: Minee Verma and Eesha Jain YouTube Channel: https://www.youtube.com/@fromstartuptoexitpodcast






