Sam Hogan is the founder and CEO of Inference, an AI infrastructure company building a private, cheaper alternative to frontier models from companies like OpenAI, Anthropic, and Google.

Instead of forcing companies to send sensitive data to third-party model providers, Inference helps businesses train and host large language models with more control, privacy, and flexibility.

In this episode, Sam explains why companies may not want to send their data to frontier AI labs, why he believes open and private model infrastructure will matter, and why Inference is trying to make large language models cheaper and easier for enterprises to run.

He also talks about raising $11 million, building one of the largest decentralized GPU clusters on Solana, why the company moved away from crowdsourced compute, and what it takes to build trus ...