Open Models Are Collapsing The Cost Of AI
In this episode, Ollama Co-founder Jeffrey Morgan explores the significant shift towards open models within enterprise settings. He highlights a 150X growth in token usage since the beginning of the year, largely attributed to coding agents. Morgan also notes that Chinese models currently dominate cloud token consumption. The discussion delves into the role of fine-tuning, comparing open and closed models, and the dynamics of local versus cloud model deployment. Additionally, he touches upon NVIDIA's involvement in the open-source ecosystem and the challenges of acquiring GPUs.
Original sources
YouTube Y Combinator
AI Radar wrote this summary from the publisher’s notes with AI, not from a transcript. How it works.