The technology behind ChatGPT is designed to give you answers that will appear to be meaningful. They are not. It is designed to tell you what it thinks you want to hear.
This is an implementation of GPT-3.5 with a reinforcement learning model. That’s it. Don’t overthink it.
All things being equal, it’s better for this kind of technology to pursue cautiously on any statements about race, gender, sexual orientation, etc.
Follow
@briannawu the sheer quantity of folks not realizing they are talking to a machine is sometimes quite jarring
