The University of Texas at Austin's Good Systems program is pleased to announce advancements in its research on critical cross-cutting themes in ethical AI. Last year, Good Systems, a leader in interdisciplinary research on the ethical development and use of artificial intelligence (AI), established three themes to bridge connections among its six core research projects and identify new lines of inquiry in the field.
Research on one of these themes, AI Alignment and Safety, began last year under the leadership of Brad Knox, a research associate professor in the Department of Computer Science. This fall, Good Systems is expanding its efforts by welcoming two postdoctoral fellows to lead research on the other two themes. Lingyuan Li will focus on Work at the Human-AI Frontier, while Jared Jensen will explore Knowledge, Generative AI and Power. These themes are designed to unify and expand upon the existing research within Good Systems, with the ultimate goal of fostering collaboration across its core projects.
Work at the Human-AI Frontier
Lingyuan Li will lead the Work at the Human-AI Frontier theme, exploring the role of AI in the future workplace, including topics such as safe and equitable skilled trade work, labor market dynamics, urban AI deployments and human values of work as they relate to AI.
With a Ph.D. in human-centered computing from Clemson University, Li brings a rich background in human-computer interaction and computer-supported cooperative work. Her doctoral research focused on how emerging technologies like AI impact social dynamics in the workplace.
“The Good Systems program offers a unique opportunity to explore the ethical implications of AI by balancing technological advancement with social responsibility,” Li said. “I’m particularly passionate about designing AI to benefit society while minimizing harm.”
Li will be a postdoctoral fellow at the School of Information and supervised by iSchool professor and Good Systems Executive Team founding chair Kenneth Fleischmann. "Dr. Li’s interdisciplinary background, also including degrees in computer science and software engineering, makes her a perfect fit for this role, which will involve building interdisciplinary collaborations across multiple core research projects,” Fleischmann said.
Before her Ph.D., Li earned an M.S. in computer science from NYU and a bachelor’s in software engineering from Southeast University in China.
Knowledge, Generative AI and Power
Jared Jensen will lead the Knowledge, Generative AI and Power theme, examining the interplay between generative AI, knowledge production and power dynamics. Jensen recently completed his Ph.D. in communication studies at UT Austin, where his research focused on the ethics and politics of technological change in creative industries, particularly among independent musicians.
"One of the things I'm hoping to be able to do is connect the different areas of Good Systems in ways that they maybe haven't thought of before, or that they've been too busy to cover,” Jensen said. “It's an opportunity for a postdoc to come in and say, ‘What are the intersections between racial equity and surveillance in AI or misinformation and disinformation? Is there a project in there that can put those three pieces together? And then, How do we scale that project? How do we configure that project from it?"
As a postdoctoral fellow, Jensen will be supervised by School of Journalism and Media professor Dhiraj Murthy, who chaired the search committee that selected Jensen. “I am looking forward to working with Dr. Jensen on a study of Good Systems itself, which examines our definitions of ethical AI, as well as on some of the synergies and intersections across core research projects,” Murthy said. "We expect to publish two articles together and submit a grant during his postdoc with Good Systems."
Building on this collaborative vision, Jensen expressed his enthusiasm for the ethical approach Good Systems takes towards AI technology. “One of the things that impresses me about Good Systems is that they frame the question from what can we do with the AI technology to what should we be doing?” Jensen said. “And I think that that's really, really valuable. I'm excited to get more invested in that kind of conversation and understand the different ways that we can approach AI. I'm most excited about that."
In addition to his Ph.D., Jensen holds an M.A. in organizational communication and technology from UT Austin and a B.A. in communication studies from Portland State University.
AI Alignment and Safety
Brad Knox is leading the AI Alignment and Safety theme. Knox, a research associate professor in the Department of Computer Science, joined Good Systems a year ago to connect their work in AI ethics to the question of how to align the goals and behaviors of AI systems with human interests. “The problem of AI alignment is a broad one,” Knox said. “It connects to numerous technical disciplines and also to research on humans, including at individual, group and cultural levels.”
Knox has brought a diverse research background to Good Systems, with expertise in machine learning, human-computer interaction and computational models of human behavior. His work focuses on the human aspects of reinforcement learning, particularly how humans can specify aligned reward functions. Knox's dissertation on human-in-the-loop reinforcement learning won several awards, and he has since contributed to various fields, including interactive character creation and toy robotics.
“Brad’s work on alignment is internationally recognized, so he was the obvious person here at UT to lead this cross-cutting theme,” said computer science professor and Good Systems Executive Team member Peter Stone, who had Knox as a Ph.D. student and has continued collaborating with him over the years. “Brad has a very clear vision for his research and is meticulous about examining scientific questions from all angles. He holds himself to a very high standard and as a result tends to produce research results that are very interesting and immediately impactful.”
According to Knox, the increasing influence of artificial intelligence necessitates a corresponding expansion of human involvement in its development process. “Good Systems serves as a model for facilitating such interdisciplinary work on AI, and it has practically provided such facilitation for me,” he said.
Knox earned his master’s and Ph.D. in computer science from UT Austin and a B.S. in psychology from Texas A&M.