It is because Substantially of the data used to train AI comes from general public substance, which include social media posts or comments, which can mirror biases like sexism or racism. A rise in large language versions or LLMs, for instance OpenAI’s ChatGPT, produces a massive tr