Fair Training in Generative AI: Towards an Ecosystem That Values Training Data Providers
450 Jane Stanford Way, Building 160, Stanford, CA 94305
Room 433A
Join us for the next CESTA lunch seminar with Ed Newton-Rex on generative AI.
Talk Abstract
Most generative AI companies use training data that has been scraped from the internet, without permission from or payment to rights holders in the works being trained on. This has led to 40 lawsuits in the U.S., and protests from creators around the world. This talk details how this has come about, the different ethical and legal opinions in the debate, and a potential path forward for AI companies and creators.
About the Speaker
Ed Newton-Rex is a composer, former tech executive, and campaigner for creators’ rights. In 2010 he founded Jukedeck, one of the first AI music generation startups, which was acquired by ByteDance. At ByteDance, Ed led the AI Music lab, then led Product for TikTok in Europe. He later led the Audio team at Stability AI, where his team’s product, Stable Audio, was named one of TIME Magazine’s best inventions of the year in 2023. In 2024 he founded Fairly Trained, a nonprofit that certifies generative AI companies for fairer training data practices. He organized the Statement on AI Training, an open letter rejecting unfair generative AI training practices signed by almost 50,000 people, and Is This What We Want?, a protest album co-written by 1,000 UK musicians. He is also a composer of choral music.
