Kind of, yeah. LLMs are hyperoptimized for one-shotting. But all of their code is like that. It's poison for the long term health of the codebase.