How cats.txt showed llms.txt evidence is GEO astrology (markwilliamscook.substack.com)

In a satirical response to the questionable evidence supporting the llms.txt standard in AI, a developer created “cats.txt,” a whimsical text file detailing fictional office cats and their roles, purr levels, and breeds. By running cats.txt through the same tests cited for llms.txt—crawling by bots, indexing by Google, and validation from ChatGPT—the creator demonstrated that these supposed proofs do not substantiate the effectiveness of llms.txt, which lacks any documented use by leading AI providers. The project sparked engagement within the SEO community, turning a joke into an unintended standard that revealed the absurdity of the evidence being used in AI and SEO practices. This initiative underscores a crucial observation in AI: the current reasoning around technological standards often accepts superficial measures as proof of efficacy without thorough validation. As the industry races to adopt new GEO (Geographic Optimization) tactics, relying on weak evidence can lead to wasted resources and misguided strategies. The cats.txt movement serves as a reminder for the AI/ML community to demand rigorous standards and verifiable outcomes, rather than succumbing to the allure of confidence in machines that reflect popular consensus rather than factual ownership.