Evaluating GPT-4's Effectiveness in Classifying Harmful Speech

How good are LLMs at classifying harmful speech? I decided to use GPT-4 to classify "misogynistic" and "homophobic" parts of speech on Kansas City Chiefs kicker Harrison Butker controversial commencement address. The results are in! https://arize.com/blog/using-generative-ai-to-evaluate-bias-in-speeches/