Are you tired of your large language model giving you boring, accurate answers? Do you sometimes wonder whether you are really getting the most hallucination out of the world’s most advanced technology?
You have come to the right place!
In this talk, you will learn practical, field-tested techniques for making highly capable language models confidently tell you complete bullshit. We will explore how to ask impossible questions, remove sensible ways out, demand certainty where none exists and generally make excellent epistemic behaviour as difficult as possible.
And the best part? You may already be using some of these techniques without even knowing it!
By the end of this session, you too should be able to take some of the world’s most advanced technology and make it confidently tell you things that are simply not true.
It is easier than you think!
