Explore interactive data visualization as a tool for studying user interface design and human-computer interaction in this 28-minute conference talk from the Next in Science series at Harvard University's Radcliffe Institute. Delve into Arvind Satyanarayan's research, which has garnered awards and is utilized by Wikipedia and the Jupyter/Python data science communities. Learn about human-centered AI, cross-filtering, declarative specification, and visualization tools like Vega-Lite, Voyager, and Lyra. Discover innovative concepts such as semantic dictionaries, neural networks, and semantic painting applications. Gain insights into domain-specific abstractions and future directions in the field of data visualization and intelligence augmentation.