Adrian JohnsonTLDR:
- University of Chicago professor, Chenhao Tan, discussed the concept of human-centered AI and its potential impact on decision making during a Technology and Social Behavior Colloquium.
- Tan emphasized the importance of incorporating human psychology, goals, and values into the development of AI systems to create a more helpful and intuitive experience for users.
University of Chicago professor, Chenhao Tan, recently spoke at the Technology and Social Behavior Colloquium, where he discussed the concept of human-centered AI and its potential impact on decision making. Tan’s research at the Chicago Human+AI Lab focuses on using AI as a complement to human decision making, rather than replacing it. He believes that AI should be built with the goal of helping humans, taking into account human psychology, goals, and values throughout the process.
In his talk, Tan highlighted the connection between human decision making and the predictive process of AI. He explained that decisions can be seen as prediction problems, using the example of cancer diagnosis. Doctors make predictions about whether a medical scan shows cancer based on their past experience. This is similar to how AI models are trained to make decisions based on data analysis. Understanding the algorithmic rationale, or “explanations,” behind AI models can help identify their faults and improve decision making.
Tan stressed the importance of incorporating knowledge of human intuition into these explanations. He believes that there is not enough focus on understanding humans, which is necessary for building effective human-centered AI systems. His interdisciplinary approach to human-centered AI, which combines computer science and communication, was highlighted as a valuable perspective in the field.
The talk resonated with attendees, including Sachita Nishal, a fourth-year Ph.D. candidate in technology and social behavior. Nishal works on designing AI systems for journalists and found Tan’s emphasis on relying on human intuition when generating explanations to be particularly important. She believes that understanding human intuition is crucial when designing AI systems for specific professions, such as journalism.
The Technology and Social Behavior Ph.D. program, which hosted the winter colloquium, focuses on exploring the impact of technology in social contexts. Tan’s talk aligns with the program’s goals and provided attendees with valuable insights into the intersection of human decision making and AI.
The field of human-centered AI is gaining increased attention as researchers and developers recognize the need for AI systems that are intuitive, helpful, and aligned with human values. Tan’s talk at the Technology and Social Behavior Colloquium provided valuable insights into this emerging field and sparked discussions about the future of AI and decision making.