147. Large language models (LLMs) - interview with Dr Guy Emerson
Education Bookcast
00:00
How Chatbots Are Trained to Be More Polite and Kind
Currently available systems have been fine-tuned to be more polite and more kind. You don't want the chatbot to keep like disagreeing with you right from it. It's very easy to underestimate how much you're doing yourself in terms of interpreting language so when you get the output of the system you think oh actually i've understood something now. There's a huge amount of what you're doing perhaps unconsciously that the system may not be able to do.
Transcript
Play full episode
