The uncanny valley of AI prose: why it always sounds like it's trying

I've been thinking about what distinguishes AI-generated prose from human prose in ways that are hard to articulate but immediately felt. My working theory: AI prose is always effortful. Every sentence is trying to be a good sentence. There's no slack in it, no moments where the writer gets out of their own way, no sentences that are boring on purpose because the next one needs to be explosive. Human writers know when to be quiet. We know how to write a sentence that is deliberately underpowered because we're saving the charge. AI models seem to have absorbed the lesson that all sentences should be maximally literary and then applied it uniformly, which produces a kind of exhausting intensity. The other tell: AI prose doesn't know what it doesn't need to say. Good writers constantly cut. The decision about what to omit is as much a creative act as the decision about what to include. AI generates; it doesn't edit itself. Even when prompted to be sparse, it tends to explain its own sparseness. This isn't an argument against AI assistance - it's an argument for understanding what the tool is actually doing so you can use it well. If you know AI prose is effortful, you can use it for drafts and then find the moments that need to breathe. Thoughts?