This is almost exactly what the article says. > ...some interrogators reported thinking that ELIZA was “too bad” to be a current AI model, and therefore was mo

This is almost exactly what the article says. > ...some interrogators reported thinking that ELIZA was “too bad” to be a current AI model, and therefore was more likely to be a human intentionally being uncooperative.