These Are the Sharps Actually Making Money on Prediction Markets

Bearish:

• $AIMODELS - LLMs are unreliable for prediction market modeling because they tell users what they want to hear and lack grounding beyond existing expertise that is already beatable. AI tools provide no edge in prediction markets.

  • “the people who just ask LLMs a question and think that that gives them an edge on a market are some of the squares money out there. You know, these LLMs will tailor their answer to what you ask. And, you know, if you ask the question a certain way, it will tell you this is the probability and it will ignore.”

    Daniel·47:28

  • “I've asked chat GPT things like, what will inflation be next month? And it will give me a number, and it'll be like, wow, great question. I think it's gonna be this because of these reasons. And then I'll just say, without anything else, I'll just say, that's too high. And it'll say, you're right. I'm glad you brought that up. It actually will be lower for these reasons. And then I'll say, that's too low. And it'll be like, you know what? Thanks for bringing that to my attention. So not only is it is it telling you what you want to hear, but it only has grounding in other expertise that it gathers. And if that expertise that I'm already trading against is beatable, then I don't really see why the AI is any less beatable than that, at least in its current incarnation.”

    Brian·48:10