IE 11 Not Supported

For optimal browsing, we recommend Chrome, Firefox or Safari browsers.

UAlbany AI Panel Optimistic About 'Empire AI' for Academia

A panel of researchers hosted by the University at Albany last week said they hope the planned "Empire AI consortium" will create a resource for both universities and private companies to purchase and share AI computing.

Grey stone sign that says University at Albany, State University of New York
(TNS) — At a panel on generative artificial intelligence hosted by the University at Albany on Jan. 31, more details emerged about Gov. Kathy Hochul's proposal to commit $275 million to establish an "Empire AI consortium" in New York.

During her State of the State address in January, Hochul said AI is the single-most consequential technological and commercial advancement since the invention of the Internet and whoever dominates the AI industry will dominate the "next era of human history."

"I'm proud to announce that New York will be the first place in the world to put that type of computing power directly in the hands of leading academic institutions that have stepped up to participate: Cornell, NYU, Columbia, RPI, and our entire SUNY and CUNY systems," the governor said. "We have geniuses at these schools ready to innovate and launch companies. Now, they'll have the power to change the world. In order to win the race for the future, we need this specific hardware. That's why I'm proposing the Empire AI Consortium to purchase and share AI computing power right here in New York."

Hochul said the state had secured more than $125 million from philanthropic and university partners and, over the next decade, the state will commit up to $275 million.

While speaking to reporters on Jan. 30, Hochul reiterated the support from universities and private companies.

"The most excited people about this are from the SUNY and CUNY schools, as well as the privates, who are saying, 'We want to be at the leading edge of this revolution. We want our researchers and our professors and our students to be able to take advantage," Hochul said. "So, we want to move quickly on that. So, we will work out all the details, but it's not complicated to me. This is about seizing the technologies of the future."

Panelists weighed in on Hochul's plan during the discussion last Wednesday night, hosted by UAlbany's College of Emergency Preparedness, Homeland Security and Cybersecurity.

James Hendler, an AI researcher at Rensselaer Polytechnic Institute, said the original intent of what came to be called Empire AI was to have a shared resource run largely by academia — rather than the technology industry. Companies could also pay the state to use the AI technology.

"So the question is, can New York provide this kind of computing power in a way that it will be cheaper to use for all of us," Hendler said. "The feeling is, if we can get that kind of supercomputing power into the state, coupled with what's happening around microelectronics — SUNY Albany is one of the leaders in several of these large initiatives — billions of dollars are going to be invested in microchips and things."

Hendler added, as of right now, a large part of that industry are AI microchips, which are "very, very expensive," and "tremendously power-consuming."

Brian Nussbaum, associate professor at UAlbany's College of Emergency Preparedness, Homeland Security and Cybersecurity, said he believes AI technology could also assist people when they interact with the government.

"AI has the potential to really help people fill out forms more efficiently and that would free up people to help with the most complicated cases, where the simpler cases can be sort of streamlined and done more effectively. So, there are places where I think it has the potential to be very useful," Nussbaum said. "One thing government has a lot of is filling out forms, filling those forms out in triplicate and submitting them to four different people ... so, I do think there is the potential for a co-pilot in filling out those sorts of forms, which could be really useful."

Hendler added that a multi-page document was also released around the State of the State address that provides guidance on how to use AI technology in state government.

"If a state government wants to deploy a chatbot on its site, it is now supposed to tell you, 'Hi, I'm an AI Chatbot,' and if it hands you off to a human, it's supposed to say, 'Hey, I can't handle that anymore. I'm handing it off to a human,'" Hendler added. "They're saying the principle is you should know if you're talking to a human or a machine, and you know, that's a great example."

Panelists also discussed some of the concerns and potential dangers around AI technology. Following the New Hampshire primary election last month, the state's attorney general is currently investigating reports of a robocall that used AI to mimic President Joe Biden's voice to discourage voters from voting.

"On the one hand, you've got to be ignorant enough in order to have it be actually useful to ask this [AI technology] a question, but you also have to be knowledgeable enough to be able to discern whether what it's giving you is reliable," Jason D'Cruz, associate professor and director of undergraduate studies in UAlbany's Department of Philosophy, said. "And one of the problems is that it sounds so fluent and the things that it says sound so plausible that a lot of those signals of expertise that we often rely on, we can no longer rely on when we are engaging with the tool."

©2024 The Daily Gazette, Schenectady, N.Y. Distributed by Tribune Content Agency, LLC.