The Reaction to Your Paper From Machine Learning Researchers

I think we are seeing lately some decent models for things like music generation. So maybe it's starting to change. And so in your paper, the most pressing problem is in language. Yes. I would say so particularly if we are only, if we're assuming that we will only use very high quality data. Yeah. It doesn't take into account all the possible improvements in data efficiency that you mentioned. Right? Can we build models that rely on less data or smaller models or models that are hybrid that incorporate knowledge graphs or information retrieval and things like this,.Yeah. The paper assumes that basically the path we're on is the path will be. yeah. Let's see.