Wrap Up

Another student not in this class has decided to try and make their own chatbot that can discuss and recommend movies. What is one piece of bias-related advice you would you give them when curating their training data?