Current Generative AI
Generative AI focuses on creating new content—text, images, 3D models, etc.—from training data using computational techniques like diffusion models. Current research emphasizes improving model efficiency (e.g., reducing computational costs for data attribution), addressing fairness concerns in generated outputs across various contexts, and resolving copyright issues through economic frameworks that fairly compensate data providers. These advancements are significantly impacting various fields, from improving the efficiency of architectural design to raising crucial ethical considerations in the development and deployment of these powerful tools.
Papers
November 9, 2024
June 16, 2024
April 25, 2024
April 22, 2024
December 18, 2023
September 13, 2023