How to detect AI-generated text
As part of my collaboration with Imaginary Cloud, I had the opportunity to present a workshop titled A Watermark for Large Language Models (based on this paper), describing and discussing the need for detection that arises from the increasing sophistication of AI text generation models. During the presentation, I focused on watermarking as a potential solution for this issue, while also touching upon several other approaches. You can watch the discussion here.
In addition to that, I co-wrote a blog post on this subject. The post delves into the various tools that have been proposed to detect AI-generated text and explores the benefits and limitations of each approach. Please click here to read the post.