AI-SDM is one of seven AI research institutes. Awarded today by the National Science Foundation(opens in new window)A five-year, $20 million commitment from the NSF will support the Institute.
“The best applications of artificial intelligence in the social realm will come when we can not only advance AI for decision-making, but also better understand human decision-making and bring the two together. ” Artie Singh(opens in new window),Professor Department of Machine Learning(opens in new window) CMU’s Faculty of Computer Science(opens in new window), serves as the director of the Institute. “Social scientists study human behavior. Machine learning researchers develop new AI technologies to aid decision-making. To maximize the impact of these technologies, social scientists and AI researchers need to work together to come up with solutions that leverage AI capabilities while ensuring social acceptance.”
AI-SDM has a computer science department and Dietrich University of Humanities and Social Sciences(opens in new window) In addition to CMU, he has conducted research at Harvard University, Children’s Hospital Boston, Howard University, Pennsylvania State University, Texas A&M University, University of Washington, MITER Corporation, Navajo Institute of Technology, and Winchester Thurston School. This diverse group of researchers and practitioners works with public health departments, emergency management agencies, non-profit organizations, businesses, hospitals and clinics to enhance decision-making.
“As artificial intelligence advances at a dizzying pace, our future lies in enabling researchers, social scientists, decision makers, and the general public to work together to understand and ethically use these tools. It’s up to you to do’ including CMU. “I am proud to present a $20 million research award from the National Science Foundation to CMU, which leads the US Institute of AI for Social Decision Making, including my alma mater, Howard University. Collaborating with several institutions, the institute works multidisciplinarily to design ethical, human-centered AI tools to improve disaster response and assist public health officials, community workers, and clinics. “
AI-SDM is the fifth NSF-funded AI institution to include researchers from CMU, and the first to be led by university expertise. CMU faculty members are the AI Lab for Cooperative Support and Responsive Interaction of Networked Groups (AI-CARING), the AI Lab for Future Edge Networks and Distributed Intelligence (AI-EDGE), Already contributing to the USDA-NIFA AI Institute Resilient Agriculture (AIIRA). ) and the Agricultural AI Institute (AgAID) for transforming the workforce and decision support. these institutions(opens in new window) Established in 2021.
“The National AI Lab is a key component of our country’s AI innovation, infrastructure, technology, education and partnership ecosystem,” said NSF Director Sethuraman Panchanathan. “These laboratories are driving discoveries that will ensure our country is at the forefront of the global AI revolution.”
By connecting AI and social science researchers, AI-SDM enables data-driven, robust and resource-efficient decision-making, addressing biases, perceptions, etc., key to accepting these decisions in the field. Improve results by taking into account different human factors. Risk, Trust, Fairness. AI-SDM aims to leverage AI to better understand human decision making. Improve AI decision-making ability. And apply those advances to create better, more reliable alternatives.
“Artificial intelligence has extraordinary potential, and at this critical stage in its development, stakeholders across disciplinary boundaries are coming together to bring these generational breakthroughs to the real world. should be applied with Teresa Mayer(opens in new window), Vice President of Research at CMU. “CMU is grateful to the National Science Foundation for its partnership, which has enabled AI-SDM and its partners to advance powerful, human-centric AI solutions for challenging situations that require split-second decision-making.” You can.”
AI-SDM embarks on several fundamental thrusts to better understand human decision-making and create AI tools to assist decision-making. Cognitive and behavioral scientists develop computational models that accurately represent how and why humans make decisions in times of crisis. Predicting human choices is key to developing better AI tools and succeeding in society. This work, Creotilde Gonzales(opens in new window)Research Professor at CMU Department of Social and Decision Science(opens in new window)and Christopher Dancy, associate professor of engineering at Penn State University.
“Our work at AI-SDM contributes to the basic research needed to accurately predict human choices under conditions of uncertainty, time constraints, and temporal dynamics. , building the future of experimental and computational cognitive decision-making science, promoting fairness and impartiality through human AI complementarity,” said Gonzalez, co-director of research at the Institute. .
Armed with this understanding, social scientists and AI researchers will work together to understand human-AI complementarity and create group and hybrid human-AI decision-making models. This also creates an understanding of how social values such as fairness, ethics and risk influence individual and group choices. Leading the work is Ariel Procaccia, Professor of Computer Science at Harvard University. Aditya Ramdas(opens in new window)Assistant Professor at CMU Department of Statistics and Data Science(opens in new window) Machine learning department.
“When AI or humans predict how a particular situation will develop, or suggest different options to take due to differing underlying perceptions of risks and benefits, these complex preferences are taken into consideration. It’s important to think about how best to draw out and combine them into the group’s decisions,” Ramdas said. “In an environment where these agents make repeated decisions, we want to design algorithms that can learn from experience how to combine decisions from AI or humans, perhaps to give different individual incentives towards a common group goal. I have.”
Finally, the Institute’s AI researchers will develop tools that can make autonomous decisions to support people in both disaster management and public health management. They have to work under intense pressure and constraints in a dynamic and uncertain environment. You have to juggle competing objectives with imperfect information and use imperfect communication to coordinate with many people. This is a big challenge for AI today. The work was led by Sham Kakade, his professor of computer science and statistics at Harvard University. Jeff Schneider(opens in new window)Research Professor at CMU Robotics Institute(opens in new window).
“We are particularly excited about the opportunity for AI to assist real people in scenarios where good decision-making is most needed but difficult to achieve. We have already developed the system, which will help support social issues.”
The AI tools created by AI-SDM not only help decision makers with the task at hand, but also help them reflect on past actions and evaluate decisions that weren’t made. Would the outcome have been different if emergency managers or public health authorities had sent resources or targeted interventions to one location rather than another? Tools that can model or simulate these scenarios would It helps you make better decisions. Extensive research in the humanities has examined how counterfactual and causal reasoning affect human decision-making and acceptance, and applying this research is key to trustworthy and explainable AI. becomes. Kun Chan(opens in new window)Associate Professor at CMU Department of Philosophy(opens in new window)will lead this effort.
“Twenty years ago, CMU helped create the modern field of causality discovery,” said Zhang. “We are now moving to a higher level to find hidden causal variables and relationships for causal and counterfactual inference from video, image, text and tabular data. Our efforts will have a direct impact not only on decision-making, but also on related areas such as scientific discovery, healthcare and marketing.”
