Just try it then.
I upvoted it because it aligns with my own findings working on real projects. Especially the bits about needing to “ground” the LLM in appropriate context, and being mindful of the sliding context window.
I upvoted it because it aligns with my own findings working on real projects. Especially the bits about needing to “ground” the LLM in appropriate context, and being mindful of the sliding context window.
Of course, I can try it. But trying it does not prove anything. It must be tested. Testing is a much higher standard than "trying" and a lot harder to do.