Enrico Schaefer breaks down when to use GPT-4o versus O3 and O3 Mini. GPT-4o is fast for drafting and creative work, while O3 uses chain-of-thought reasoning to break problems into pieces and evaluate answers in parallel, cutting hallucinations. You can switch models mid-conversation, though the feature isn't yet in ChatGPT projects or assistants.
Chapters
Welcome to AI Winning Hacks for Business Professionals. This is my series of short video tips and AI hacks that will give you a big edge in your business workflow. Today's winning AI hack is brought to you by AI for the Win, which is the largest community of business professionals using AI to win at work, and also the Travis AI YouTube channel. This is your go-to hub for for cutting-edge AI strategies and insights to stay ahead of your competitors in the AI age. Remember, one small change in your AI game can result in big results. My name is Enrico Schaefer, and I'm here to help you harness the power of secure, ready-for-prime-time AI tools and techniques. All the links you need to succeed are in the description below. Now let's dive into today's game-winning AI hack for business professionals. For today's Power Tip, I want to talk a little bit about how you can navigate between the various GPT models. So you've got GPT-4o, which is a generative AI model, but you've also got 0.3 and other models that you can use. What you may not know is that you can actually switch between the models during a conversation. So let me show you how this works. First thing before we get started is so far this is not available when you use ChatGPT in projects or when you create a GPT assistant. Switching between models is spotty at best at this point, but as with everything with OpenAI, they continue to add base functionality to all the different aspects of the platform. They add it Initially in the web version, and then eventually they put it in the desktop version. So let's go ahead and take a look at how we can use the switching between models to your advantage. First things first, you need to know that Generative AI 4.0 is generating text. It is guessing the next word, and it's using the neural networks and the transformers in order to try and figure out what is the next best word to put into the system. Analytical models, which will be O3 and O3 Mini, they're doing something on top of generative AI. They are actually analyzing using chain-of-thought reasoning a particular problem. And what that means is instead of just guessing the next word really quickly, they're going to reformulate the question in lots of different ways. They're going to break the question down into component parts. They're going to ask that question, get some answers, and then after they get some answers, they're going to go back and compare it to the other aspects of the question or alternative ways of asking the questions. And it's doing this in nanoseconds. So it's doing this in something called in parallel. And in parallel means instead of having a sequence of question and answers and logical steps, it's doing almost instantaneously at the same time. So the speeds are incredible. So if you want to switch from GPT text generation to analysis or reasoning, you simply select the GPT-03 Mini model, and now you ask it to do an analysis of some problem. The distinction between these 2 different approaches, these 2 different models, and how you are going to use them is in the purpose. Are you trying to generate text or summarize information In which case you are going to use O4. If you want analysis or in-depth thinking or logic or reasoning, you want to use O3 Mini. So maneuvering between these models as you do your problem solving is mission critical. It is a necessary skill that you need in order to succeed using AI as a business professional. Understanding the difference between 4.0 generative AI And O3 analytical models, which build on top of generative AI and slow the process down, add an additional layer of attention mechanisms that will allow the model to do chain of thought or tree of thought reasoning, just like a room of scientists or lawyers might do if you had 50 in a room and they were all chipping in and commenting and testing and challenging back and forth, back and forth, back and forth. That's exactly what's happening. with these new reasoning models. And that's why they work so well. That's why hallucinations are way down. That's why their output can be so much more meaningful and comprehensive on complex issues. Entrepreneurs, CEOs, lawyers, other business professionals, we're in the business of solving problems. We're getting paid for our brain. We're getting paid for our analytical abilities, our ability to analyze and make decisions. That's That's exactly what's going on with these analytical models. So this is new layers of reasoning that's been placed on top of the GPT models. It's just starting to come of age in late 2024, now early 2025. And it is a major leap forward in what you are able to do with AI at work. So start to learn how to jump back and forth between the different models. Understand that eventually within projects, within GPTs, you'll be able to do the same thing as the functionality comes along and start to get a feel for when you use one model versus the other. My name is Enrico. We'll see you at the next Power Tip.
