Beware of Glorbo: A Use Case and Survey of the Fight Against LLMs Disseminating Misinformation
In the Summer of 2023, a website called Z League published a glowing article about a new World of Warcraft character: Glorbo. The problem? Glorbo was the invention of Reddit pranksters drawing attention to the practice of AI-generated game articles.
Amusing, but generative AIs pose an enormous threat in the fight against mis/disinformation. Things are moving quickly. How does one get their bearings?
This talk is a survey of the Trustworthy AI landscape: What does the term “Trustworthy” specifically mean? What are the main concepts in the world of Trustworthy AI? What are the proposed approaches? What has been achieved so far? How can Trustworthy AI frameworks be taken into account when your stakeholder requests generative AI for your next Information Architecture project?
This talk is for those coming to terms with how generative AI tools impact future information architecture projects processes. And for those who advocate for trustworthy information.