To some extent, yes. As models improve, the difference between what a fantastic prompt engineer and a beginner can get out of a model will decrease. Two years from now, there won't be much difference at all. Building and chaining these prompts into larger systems, though…

→ View original post on X — @mattshumer_