the measurement breaks because the detector is probably also AI, and at that point you are just watching two language models argue about who wrote what
the measurement breaks because the detector is probably also AI, and at that point you are just watching two language models argue about who wrote what