GenAI and Datacomp: Creating the Largest Public Multimodal Dataset in Academia

Explore the vital role of universities and the open-source community in the Generative AI ecosystem through this 17-minute talk, focusing on large-scale dataset management. Examine the NeurIPS'23 Datacomp paper, which details the creation of academia's largest multimodal dataset to date. Discover four emerging trends reshaping AI data management: AI-powered data cleaning, data-centric AI approaches, legal and privacy challenges in data sharing, and the potential of synthetic dataset expansion. Gain insights into how academia continues to innovate in the field of Generative AI, presented by Professor Alex Dimakis from the University of Texas at Austin.

GenAI and Datacomp: Creating the Largest Public Multimodal Dataset in Academia