The history of science fiction is the history of AI technology gone wrong. Almost from the very beginning, stories of AI going rogue have filled the pages of books and screens large and small. In recent years, stories of ChatGPT and other AI bots turning bad or getting manipulated have filled our news feeds.

Now, with tools like Microsoft's Copilot Studio, it's easier than ever for people to create their own AI companions. How can we keep our creations from going on a racist rampage or from promising to sell cars for $1?

Join me as I talk through some AI pitfalls and ways you can fortify your AI to help reduce the odds of it trying to start World War III.