Can you still tell human writing from machine output? A cognitive benchmark measuring your ability to distinguish authentic human text from AI-generated responses.
You will be presented with a question and a single answer. Your task is to determine whether the answer was written by a human or generated by ChatGPT. Both responses are real — sourced from the Human ChatGPT Comparison Corpus.
Each session draws fresh samples from a pool of over 17,000 paired responses. No two tests are identical. Your results are scored and mapped to a detection-IQ scale.