Community Driven Data Collection and Consent in AI - Jessica Rose
About this talk
This talk addresses the consent issues surrounding generative AI in 2024, particularly regarding the use of scraped and unauthorized datasets that can undermine original content creators. The speaker explores the Common Voice project, which gathers volunteer-donated speech data to foster linguistic diversity in digital products and services. By demonstrating how community-driven dataset collection and governance can support ethical AI development, the session highlights the importance of data trusts and consent-led models. Ultimately, it presents a vision for a future where freely donated data promotes a more inclusive and less dystopian AI landscape.