I think you’re correct in the sense that the original study probably intended to cast ChatGPT as racist, so published statistically insignificant findings to su

I think you’re correct in the sense that the original study probably intended to cast ChatGPT as racist, so published statistically insignificant findings to support their claim. They went in with a bias against AI in the first place, and there’s a probability they used the label of racist because it is the most efficient negative signal in educated left-leaning circles, rather than it being a natural conclusion from a standard route of scientific inquiry.