> They're useful. But I keep seeing people try to apply the same four linear methods to every problem in sight. Nowadays I see the reverse. People loading in h

> They're useful. But I keep seeing people try to apply the same four linear methods to every problem in sight. Nowadays I see the reverse. People loading in huge huggingface models, when a simple linear regression would have been fine.