The United States is at the forefront of AI research and safety guidelines, and the Luddy School of Informatics, Computing, and Engineering and IU’s Cognitive Science program are well positioned to play a big role.
In fact, said Johan Bollen, Informatics chair and professor of Informatics and Cognitive Science, they already do.
Bollen attended a White House conference on AI Aspirations in Washington D.C. to discuss the federal government’s efforts for practical artificial intelligence applications in areas such as drug development, energy grid resilience, materials development for semiconductors, education efficiency and weather models.
He was one of about two hundred people invited consisting of cabinet secretaries, members of Congress, high-ranking administration officials, program directors, AI technology leaders, scientists and researchers. It was co-organized by the National Science Foundation and the Office of Science and Technology Policy.
“What struck me was how ambitious this administration is when it comes to leveraging AI to serve the public,” Bollen says. “They’re not playing. They are very committed. They are supporting AI innovation as well as safeguarding its use through regulation and guidelines.
“That’s good news for Luddy and the Cognitive Science program; we’re making a big push in AI research and development. We’re very well positioned given our eminence in complex networks and systems, cognitive science, human-computer interactions, robotics, health, knowledge discovery, chip development, engineering, and cybersecurity.”
Senators Amy Klobuchar and Mark Warner, Congressman Don Beyer, U.S. Secretary of Education Miguel Cardona and U.S. Secretary of Commerce Gina Raimondo presented along with industry leaders and government agencies such as the Defense Advanced Research Projects Agency, which is responsible for the development of emerging technologies for military use, and Advanced Research Projects Agency-Energy, which promotes and funds research of advanced energy technologies.
Conference topics included AI for Weather Prediction, Accelerating the Promise of AI for New Therapeutics, A Decarbonized and Reliable Electric Grid Enabled by AI, Government Services Right When They Are Needed, and AI for Transportation Safety Projects.
“Much of what was discussed, the concerns they had and the potential they see in AI is aligned with the work we’re doing at IU,” Bollen says.
Bollen says companies such as OpenAI, Microsoft, IBM, Google and Anthropic are pushing AI development, but regulation and safeguarding are needed. To address this, the government has formed the U.S. Artificial Intelligence Safety Institute.
“In the United States,” Bollen says, “I have the impression that the general attitude is look at the amazing things we can do with AI. Let’s talk about what we can do, but also work toward making it safe and equitable.
“In Europe, they seem a bit more concerned about getting it right from the get-go from the perspective of guidelines and regulations.”
The senators stressed the importance of regulations, using social media as an example. It took off without much regulation and the result, Bollen says, were “wild-west” consequences regarding privacy, health and mental health.
“The consequences of social media were under-estimated at the time,” he says. “The senators said we should try to learn a lesson from that.”
Discussions included using AI to develop drugs in years rather than decades. AI could improve the entire chain of drug discovery, including getting government approval.
“We don’t know if it’s possible,” Bollen says, “but it’s exciting stuff. To see it discussed by people at that level of government was amazing. It was a daring and compelling vision of the future.”
Presenters talked about improving weather forecasting as well as the safety and capacity of the transportation system; they addressed accelerating the development of semiconductor chips crucial to AI training, which is something IU and Luddy are doing, Bollen says.
A representative of the Office of Management and Budget discussed how AI can improve access to government services when they are needed. The vision that Bollen heard was that AI could make resources easier to understand, more accessible, and more pro-active. Bollen says if a natural disaster damages your home, you might need to contact FEMA or other government services, and provide the proper documentation, which can be confusing and difficult. AI can pro-actively expedite the process, including language translation if needed.
“AI could make services available before you even figure out who to call,” Bollen says. “It can adjust how information is delivered to your needs and abilities.”
AI can help ensure the nation’s energy grid stays reliable amid escalating demand. Looking ahead, Congressman Beyer talked about his vision for fusion, which can deliver an almost inexhaustible source of energy without producing greenhouse gases or radioactive waste, but presents unresolved engineering problems -- heating plasma to enormous temperatures through a controlled thermonuclear fusion reaction, and then containing it -- that AI could solve.
“The engineering and control systems needed to achieve fusion will have to be very sophisticated,” Bollen says. “This is where AI can really speed up development and make the old dream of cheap, clean and abundant energy a reality.”
Bollen's invitation followed a pair of papers he co-authored that appeared in Nature, the world's leading multidisciplinary science journal. The first paper, ChatGPT: five priorities for research, listed a series of recommendations on how to use it in scientific research.
The second paper, Living guidelines for generative AI – why scientists much oversee its use, addressed establishing an independent body to test and certify generative artificial intelligence before the technology damages science and public trust. Bollen said UNESCO and the Organization for Economic Co-operation and Development advised on the development of their living guidelines for the applications of AI in science. The European Research Council of the European Union later issued guidelines similar to their recommendations.
“We proposed guidelines that could safeguard society and science from inappropriate applications of AI technology,” Bollen says, “and make sure we benefit from this technology because the potential is enormous.
“From what I learned at the conference, this administration sees it that way, as well. I was impressed in how far evolved our government is in their thinking about using AI to improve the lives of Americans and people across the world.
“These were bold dreams and aspirations. They were fantastic ideas. I don’t know if all of it is possible, but they’re great aspirations to which Luddy and the Cognitive Science program can make important contributions.”
Bollen says the work will take years and involve multiyear projects featuring multiple research and development efforts. Luddy and the Cognitive Science program will play prominent roles as they continue to develop their capabilities as part of the brain-machine collaborative space.