Why Flash Models, Not Frontier Models, Will Win in 2026

In this monologue podcast, Nate B Jones shares his optimistic outlook for artificial intelligence in 2026, highlighting a shift from judging AI by hype to judging it by practical results. He reflects on the bursting of the AI bubble in 2025 and the importance of focusing on real-world applications, protocols, and constraints. Nate expresses excitement for agentic workflows, understanding entropy in LLM systems, and the emergence of a post-ChatGPT software future, including graphical AI and robotics. He also notes the repricing of careers around dual fluency, where individuals possess both AI expertise and domain-specific knowledge.