The Biggest Limitations of Mat Learning Research

In practice, i could envision a world in which your forecasts are very good. They help and form the decisions about inventory, and the inventory is all changed. So it's a dynamic system. There's a rue. I'm not sure if you're familiar with it, but i'll share in case some listeners are not. It's called goodhart's law. And goodheart's law states that as soon as any metric becomes important, it ceases to be a good metric,. Almost that having it kind of ruins it.