Hacker News Reader

Is there any way these LLM tools watermark their output in a way that keeps them from re-training on output generated from the same LLM?

strategic em-dash placement is my guess but only the machine knows the code

Nice try, AI!