AI already is part of our everyday lives and will inevitably continue to grow. But one important question lingers: Is AI, particularly the large language models that have exploded since ChatGPT appeared in 2022, actually good for people?
Joe Yun, research professor of electrical and computer engineering in the Swanson School of Engineering, has an answer: “Because it is inevitable, we have to make AI as good as we can for everyone. Obviously, there is a lot that could go very badly. But we need to work for the good so that it doesn’t end up that way.”
Yun also is Pitt’s director of artificial intelligence enablement, a new role in which he leads efforts to define the roles of AI at the University to help students, faculty and staff adjust to and learn the ways they can use AI technology. He is leading a project to build an internal generative AI model tailored specifically for Pitt and connecting University data systems.
“We envision a future where faculty, staff and students could ask generative AI-type questions in a system that brings together different information sources,” he says.
“We envision a future where faculty, staff and students could ask generative AI-type questions in a system that brings together different information sources,” he says.
Yun predicts that mature generative AI will democratize software, essentially reviving the excitement and do-it-yourself vibe of computing in the 1980s. “We’re growing that capability at Pitt for college students,” he says. “AI is not like just using the Internet. It is about creating something.”