The Importance of Causal Inference in Machine Learning

The most basic use case for causal modeling is to be able to predict the outcomes of interventions. If you are affecting the data generating process, then you're changing it. And so if you have a causal model, your causal model can account for that change and its predictions will still be valid. But as we know with even some of the more powerful models of machine learning, they don't have a causal structure to them. So once we can predict interventions, we can do a lot of clever things on top of that. For example, we could have models or modules that are more robust and can transfer across domains. That's really the main thing when I pitch people on like, why should