Poisoning Well — fighting back against LLMs by creating deliberately nonsensical versions of his articles hoping to subtly poison the AI models'