Updated pre-print (http://arxiv.org/abs/2308.01264) testing moral and legal intuitions in GPT-4, Claude 2.1, Llama 2, and Gemini Pro across a series of different studies. We found that...

Comments