Artificial intelligence is a key focus of the University of Wisconsin-Madison’s Research, Innovation and Scholarly Excellence (RISE) initiative and the idea — the Wisconsin Idea — is to supercharge what the university is already doing in the AI space to benefit business and society.
Chancellor Jennifer Mnookin launched the initiative, which also will concentrate on health and sustainability, in February 2024. RISE AI is the first component of the initiative, which will require accelerated faculty hiring, building research infrastructure and interdisciplinary collaboration. As the program develops, there will be opportunities for the business community to contribute as well.
UW-Madison faculty members already are conducting research with the help of the computational power of AI models. Among the advances so far are the development of an AI-driven screening tool that helps hospitals identify patients at risk for opioid use disorder and reduce costly opioid-related readmissions, and the use of AI to help identify sex-specific risks associated with brain tumors.
The aim of the Wisconsin RISE Initiative is to take AI-driven discovery to even higher levels across more campus disciplines. A portion of the funding will come from revenue generated by enrollment increases and revenue growth from existing UW research productivity — its research portfolio grows by an average of 8% annually — and last October, the Wisconsin Alumni Research Foundation, the university’s licensing arm, announced it will commit $15 million over the next three years in support of RISE AI.
Overall, UW-Madison expects to hire between 120-150 new faculty through the RISE initiative over the next three to five years, a 40% increase over regular faculty hiring. Part of WARF’s award will be used to attract the research talent that will discover new applications for artificial intelligence. So far, the university has hired 33 of the roughly 70 new faculty members it plans to add for RISE AI alone.
John Zumbrunnen, senior vice provost for academic affairs and vice provost for teaching and learning, leads RISE AI. He said the university already has AI expertise across campus, and the new faculty hires will be needed to grow its scholarly expertise.
“Faculty hiring takes time, and we’ve made a good start this year,” Zumbrunnen said. “We’ve known all along that it would take us some time to build up to the total number of faculty that we are looking for, and that two- or three-year buildup of faculty expertise works well for giving us the kind of on-ramp to a couple other pieces.”
One of those pieces is research computing infrastructure, including a new campus-level research computing strategy, to serve faculty working in this space and new colleagues. Zumbrunnen said AI researchers have a different level of need, and a real focus of the past year has been working to coordinate cross-campus units to provide the necessary computing infrastructure.
Collaboration is another piece, and that is being developed through campus units such as the School of Computer Data and Information Sciences (CDIS), the Data Science Institute and other campus research centers that conduct automated data analysis or AI work. The Data Science Institute is directed by physics professor Kyle Cranmer and already has more than 50 faculty affiliates across campus, all of whom are working in data science and engaged with AI.
“We already have networks of people,” Zumbrunnen said, “and what we’re working to do is connect with those networks and look for where we can enhance pathways to collaboration across disciplines so that when we welcome the new colleagues we’re hiring through RISE, they come into an already enriched collaborative space.”
Cranmer’s main research interest is the influence of machine learning on the physical sciences, but as the director of the Data Science Institute, he’s trying to support all of the RISE AI initiatives.
“Instead of it just being a faculty member together with some graduate students and a postdoc, the really impactful things often require some additional technical expertise and some engineering skill, especially if you want to produce tools that people will use outside of academia that have real impact,” Cranmer said.
The Data Science Institute will hire several machine learning engineers that can partner with researchers to help execute their work.
“That’s pretty much the model that you see in industry,” said Cranmer, who came to UW-Madison from New York University.
Cranmer said future AI discoveries will go far beyond ideating with large language model (LLM) tools such as ChatGPT and will affect everything from agriculture to meteorology to energy.
“With the advances that are happening now, AI is just getting more and more powerful,” Cranmer said. “There’s some really exciting work around fusion reactors. You have these big fusion reactors and you’re trying to control hot plasma, and you can use the traditional approaches that engineers have built for controlling something like that, but what we’re seeing now is that some of these AI models can approximate these very complicated systems.”
Industry partners
When the university invokes the Wisconsin Idea, it is referring to its guiding principle that the university’s work should benefit people around the state and beyond, including the business community. As the UW builds RISE AI’s collaborative structures, which it will do with help from its Office of Business Engagement and the Schools of Business and Engineering, Zumbrunnen said a fuller picture of the private sector’s opportunities for collaboration will emerge.
The business operational benefits of AI are growing by the day, as the technology already is embedded in a variety of software systems and more applications are being explored.
“The dean of our School of Veterinary Medicine and I were chatting, and he was talking about how AI is going to impact the practice of veterinary medicine on the health side,” Zumbrunnen said. “AI is going to end up helping radiologists more quickly and accurately process and read imaging.
“When you think about a small veterinary practice, AI’s going to impact their business operations, too. It’s going to impact their financial work and their HR work, and one of the really interesting things is this isn’t going to look like a small business person going to ChatGPT and asking it questions. What we’re increasingly going to find is that AI-powered capabilities are just going to be present in the software that comes into our business spaces.”
Cranmer said AI’s benefits will extend to sustainability as well.
“There’s a lot of startups and business opportunities around different materials for batteries,” he said. “Carbon capture and carbon sequestration is another active area, but there are materials for all sorts of different things where AI can accelerate that process.”
Across disciplines
Zumbrunnen said the reaction of UW faculty to RISE AI has been enthusiastic because virtually every discipline on campus will be involved.
“AI is not just a computer science thing — it’s in really every discipline around campus,” Zumbrunnen said. “Our faculty and staff know that and they’re excited about the university’s determination to supercharge that collaboration around campus.”
