The opposite is true for me. Using LLMs to write things that I can easily write by hand is the most fun use case for LLMs. It's when I use them to write things that would be difficult to write by hand (e.g. for me, a video encoder, compiler, or game engine) that the fun is stolen.