Fall Research Exchange series tackled AI framing, learning models and safety

Illustration of a white robot hand touching a light blue network.

In fall 2023, the CITRIS and the Banatao Institute, alongside the UC Berkeley College of Computing, Data Science and Society (CDSS) and the Berkeley Artificial Intelligence Research (BAIR) Laboratory, hosted three Distinguished Lectures on the Status and Future of AI, continuing the momentum of a successful spring series.

As part of the CITRIS Research Exchange seminars, these talks featured thought leaders across disciplines working to challenge and expand perspectives on information technology and society.

Jaron Lanier

In his Sept. 13 lecture, Jaron Lanier, Microsoft’s prime unifying scientist, spoke on reframing AI as a technology that stems from humanity rather than a foreign entity that society must combat. In these terms, he argues that we can effectively integrate AI as a tool that does work for society, not vice versa, and make progress on solutions to issues such as job displacement and creating an inclusive future while improving the technology.

Alison Gopnik

On Sept. 27, Alison Gopnik, Distinguished Professor of Psychology at UC Berkeley, delivered a lecture on the parallels between young children’s intelligence and that of AI’s large language models. Based on the exploratory qualities that make young children apt learners, she stresses the importance of incorporating an element of curiosity into large language models in order to induce similar learning capabilities. 

Anca Dragan

In an effort to make AI systems safer, Anca Dragan, associate professor of electrical engineering and computer sciences at UC Berkeley, explores designing robots to continually ask and learn what humans want. Her Oct. 6 lecture discussed the challenge of aligning the capability of the technology with what is beneficial to individuals.