Comment by xboxnolifes

If the underlying prompt of the model stays the same, it seems to me that LLMs will always have common tells unless overridden with a thorough prompt from the end user. It's like if you had 1 person write half of the content on the internet. You'd probably get pretty good at noticing their writing style.

Maybe my understanding of LLMs is wrong, but it seems obvious to me that when you have a large corpus of LLM output you will eventually notice common tells when everyone is using the same models, weights, and base prompt.