I thought it fairly obvious that the poster you're responding to was making an analogy about using GPT/LLM models for decision-making, not actually endorsing pl

I thought it fairly obvious that the poster you're responding to was making an analogy about using GPT/LLM models for decision-making, not actually endorsing playing the lottery. And they're correct. Don't use generative models for extractive tasks. That's like using the handle of a screwdriver as a hammer. Sure, it might work for a while, up until there's a screwdriver in your eye.