AI Model Comparison: Which AI to Use for Different Tasks (ChatGPT vs Claude vs LLaMA)

What AI should you use for what jobs? Different models are better at different tasks, but knowing which model to use for each task is kind of hard. I mean, yeah, you could run all the different models and see what works best for a particular task that you're trying to do. But here's a quick and easy cheat sheet which is going to get you in the right direction. So this table I saw comes from the excellent Ethan Mollick. Make sure you're following him on Twitter. And basically what it shows is the different scores that have been achieved by different models in the SATs and AP. Basically we're looking at Claude, GPT4O and Llama as the top three models and nobody else really gets in here. Very broadly, if you're looking at more mathematical or economic problems, you should be using LLAMA for more language based problems. It's kind of a toss up between Claude 3.5 and GPT4O by ChatGPT. So that tends to be the breakdown. If it's more literary language based, it's going to be ChatGPT or Sonnet. If it is more mathematical, then it's going to be Facebook's llama. What's probably more interesting though is just how high or of these scores are with any model. You're going to get pretty good results. Now, obviously these are exam questions. It's not going to be indicative of real life problems, but it gives us a basic idea of what DIP models are good for, what uses.