Stop using the term "Hallucinations". GPT models are not aware, do not have understanding, and are not conscious. We should refrain anthropomorphizing GPT models. GPT models sometime produce bad output. Start using the term "Bad Output".