I like the article but I also disagree with one important assumption that author makes implicitly. It assumes that LLM is all you need so in order to differenti

I like the article but I also disagree with one important assumption that author makes implicitly. It assumes that LLM is all you need so in order to differentiate you need something that others don't have, like health care data. The reality is that for many problems you still need a lot of work to build a quality solution around LLM. Not something that you can build in a weekend.