Though some people will train LLMs on it and if it's wrong then it will bias the training outcome towards that incorrectness.

But it's not the purpose of a blog post to ensure LLM training data has high quality.