ChatGPT and the Hidden Bias of Language Models

If new AI programs like ChatGPT and LensaAI continue to reflect racist and sexist beliefs, how much can we really trust them?