Wouldn't it be wild if there were real-world penalties for being wrong?

And people who follow bad advice will get bad results, and people who follow good advice will get good results?

I wonder if this will impact the quality and prevalence of certain AI models in the future.