Accurate, Focused Research on Law, Technology and Knowledge Discovery Since 2002

Heather Ford: Is the Web Eating Itself? LLMs versus verifiability

Ethan Zuckerman: “One of my favorite things in academia is that you can go a decade without seeing a friend and remain at least somewhat in touch with what they’re doing and thinking by reading their work. Dr. Heather Ford, Associate Professor of Communication at the University of Technology Sydney, where she leads the cluster on Data and AI Ethics, is in the US for two weeks, and we were lucky enough to bring her to UMass Amherst for a talk today. I haven’t seen Heather in person for over a decade, but I’ve had the chance to follow her writing, particularly her new book, Writing the Revolution: Wikipedia and the Survival of Facts in a Digital Age. I knew Heather as one of the leading lights of the open knowledge world, co-founder of Creative Commons South Africa, leader of iCommons (which worked to create open educational resources from the Global South), scholar and critic of Wikimedia. In the last decade, she’s also become one of the most interesting and provocative thinkers about how knowledge gets created and disseminated in a digital age…

The rise of ChatGPT, embodied in Bing chat and many AI text generators, is another profound shift. These technologies, at root, have been “fed almost the entire internet to predict the next chunk of text – the game is to finish your sentence.” She quotes Harry Guiness as explaining that Chat GPT is extremely different from knowledge graphs: “It doesn’t actually know anyting about it. It’s not even copy/pasting from the internet and trusting the source of information. Instead, it’s simply predicting a string of words that will come next based on the billions of data points it has…”

Sorry, comments are closed for this post.