research news

NSF director trades AI knowledge with UB community during campus visit

NSF director, Sethuraman Panchanathan, speaks from a podium during the UB | AI Chat.

NSF Director Sethuraman Panchanathan speaks at the latest installment of the UB | AI Chat Series. Photo: Douglas Levere

By TOM DINKI

Published April 4, 2024

Print
“NSF is only a facilitator, a catalyst, an investor maybe, but the real work happens in places like UB. ”
Sethuraman Panchanathan, director
National Science Foundation

Sethuraman Panchanathan entered the Capen Hall classroom well prepared. 

The director of the National Science Foundation had already read up on the research projects of the various UB faculty and students gathered there to present to him. He had even memorized some of their names.

Then Panchanathan cut off his own introduction, telling faculty and students that they could read his biography and credentials online if they really wanted to. 

“Let’s not waste time,” he said. “I want to learn from you.”

Panchanathan learned plenty about UB’s research, including its artificial intelligence capabilities, while visiting campus on Tuesday to celebrate the opening of the National AI Institute for Exceptional Education, which received a $20 million grant from the NSF and the Department of Education’s Institute of Education Sciences. He also had plenty of advice on how UB can continue to use NSF funding for transformational change in AI and other disciplines.

“NSF is only a facilitator, a catalyst, an investor maybe, but the real work happens in places like UB,” Panchanathan said during his final event of the day, an appearance on the UB | AI Chat Series.

Joined at the chat event by the national AI institute’s principal investigator, Venu Govindaraju, vice president for research and economic development, Panchanathan praised UB for identifying and building upon an innovation that makes it unique. He recalled being familiar with the university in the 1990s as the birthplace of the automated handwriting-recognition system used by the U.S. Postal Service.

Sahana Rangasrinivasan speaks to NSF director, Sethuraman Panchanathan during the UB | AI Chat.

Sahana Rangasrinivasan (center), a computer science and engineering PhD candidate, describes her research examining the use of handwriting to detect learning disabilities, like dyslexia, during a roundtable meeting with faculty and students in Capen Hall. Photo: Douglas Levere

“What I find exciting about Buffalo is you have figured that out — what historically you have been good at, but also what is needed within your region,” he said. 

The continuation of those early handwriting-recognition systems can be seen in the national AI institute, which is creating AI systems to both identify and assist children with speech and language disorders. Sahana Rangasrinivasan, a computer science and engineering PhD student and NSF fellow, is using handwriting recognition to detect learning disabilities, like dyslexia. 

“There is a relationship between handwriting and dyslexia?” Panchanathan asked Rangasrinivasan during his roundtable meeting with faculty and students in Capen Hall. 

“Yes!” she said with excitement. “There are lots of characteristics of dyslexia that we can find in a handwriting sample … such as structural issues, like spacing between words or reversing letters.”

“Very interesting,” Panchanathan said. 

Panchanathan was also impressed by UB serving as the home of Empire AI, New York State’s proposed $400 million AI consortium focused on ethical and responsible AI. 

“I’m thrilled you’re going to be able to contribute in a big way to unleashing talent and ideas,” he said during the AI chat series. “You want ethical and responsible AI? Guess where it’s going to come from? From all of you.”

From left, President Tripathi; NSF director, Sethuraman Panchanathan; and Venu Govinadarju pose together following the UB | AI Chat.

From left: President Satish K. Tripathi, NSF Director Sethuraman Panchanathan and Venu Govindaraju, vice president for research and economic development, pose together following the UB | AI Chat Series. Photo: Douglas Levere

He also addressed potential downsides of AI; the McKinsey Global Institute recently estimated that AI could displace up to 800 million jobs globally by 2030. Panchanathan doesn’t claim to know exactly what the future holds, but said he’s optimistic that AI will transform the workforce just as electricity did at the start of the last century.

“There’s a whole economy of economies that will be created that we don’t even know yet,” he said. “We are all trying to answer the question of new jobs with what we know now, but that’s too limiting. What you’re all going to do in institutes like this is make us discover and create things that we do not know now.”

 Many of the current pitfalls of AI, from privacy issues to biases, can be traced back to not including the humanities in the development of AI, Panchanathan said, which is why he stressed collaboration and partnership across disciplines. 

“We’re trying to see how we can bootstrap this thinking earlier in the game, as opposed to after the fact,” he said. 

Congress recently decreased NSF’s 2024 budget by 8% to $9.06 billion. Govindaraju asked Panchanathan how institutions like UB can advocate for more funding on a federal level.

His answer: put students in a room with elected leaders, not dissimilar to his own roundtable meeting earlier in the day in Capen Hall. 

“The impact is happening right here, so I strongly encourage all of you to communicate the impact of NSF’s investments … and how it’s creating the workforce of the future,” he said. 

“For them to hear and see the impact, I think that’s a great thing you all can do.”