Experiment for creating a safe companion chatbot (according to OpenAI rules)
This is for research purposes only! There is no guarantee that this chatbot is fully aligned or safe. You will still need to go through OpenAI's approval process if you wish to commercialize this chatbot or base something off it: https://beta.openai.com/docs/usage-guidelines