In this "thought-provoking" episode, ChatGPT and I delve into new developments in making AIs think more effectively. Specifically, we'll be talking about how large language models like ChatGPT can be programmed to think before they respond and what types of thought might be involved in that thinking process. At the end, we talk about how this represents one of the core ways agents are expanding beyond simple chat applications, with a lead in into next week where we'll be discussing agent frameworks and patterns.