AI Institute: Planning: Foundations of Intelligence in Natural and Artificial Systems

点击翻译按钮获取中文摘要

This project will develop a blueprint for a future artificial intelligence (AI) institute that will study the foundations of intelligence across a broad array of natural and technological systems. Given the pervasiveness of AI technologies in everyday life and the expected increased impact of these technologies in upcoming decades, it is imperative that AI researchers develop more reliable, general, and adaptable systems, ones that are trustworthy and that can more successfully collaborate with humans. New breakthroughs in AI research are most likely to come from concentrated collaborative efforts between AI specialists and researchers who think deeply about the nature of intelligence in different disciplines. Via focused workshops, seminars, and numerous education and outreach activities, this project will bring together experts from a broad swath of disciplines to map out ways in which insights from intelligent systems in nature can inform and inspire AI progress. The results will be disseminated via workshop reports and educational materials, and, most importantly, a detailed proposal for an institute that would advance AI through a broader and deeper view of intelligence, both natural and artificial, taking the time to formulate longer-term questions and approaches.The goals of this planning project are to (1) map out the important research themes, questions, methods, and applications that would be the topics of the prospective institute, as well as the needed infrastructure; (2) build research collaborations and perform preliminary research; (3) plan for how the institute will engage with the overall national and international AI community, and for the institute’s education and outreach programs. This two-year planning project will explore five research themes that are central to the study of intelligence and representative of the kinds of interdisciplinary problems that the envisioned AI institute would address: 1. A Taxonomy of Intelligence across Disciplines; 2. Development and Life History of Intelligent Systems; 3. Concept Formation, Abstraction, and Analogy; 4. Collective Intelligence; 5. Evolutionary and Co-Evolutionary Intelligence. These themes will be explored in numerous workshops, project meetings, and educational and outreach activities that will bring together researchers from diverse disciplines to map out questions and methods as well as to engage in preliminary research driven by these questions. The results of all of these activities will be summarized in publicly available reports, and synthesized in a detailed blueprint for an envisioned AI institute.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.