Is There a Need for More General Machine Learning Models?

When we're thinking about machine learning models in the long term, more general models have to be few shop learners. I used to work on a problem where we were mapping natural language to actions in the web browser. And this was an angle for learning to map language to imperative, actionable behavior without having to go through some sort of intermediate sematic representation. The supposed benefits of an approach like this are that it's actually easy to gather data because people know how to operate the browser.