NSF Award
2421845
The NSF-Simons AI Institute for the Sky (the SkAI Institute) will draw together top researchers across the country to unleash the discovery potential of revolutionary sky surveys and solve fundamental problems at the forefront of astrophysics and artificial intelligence (AI), transforming our understanding of stars, galaxy formation and evolution, and cosmology and the early Universe. Fulfilling this promise requires overcoming enormous challenges in analysis and inference from enormous, heterogeneous, multi-modal datasets, developing physically accurate simulations, and designing ever more complex astronomical instruments and surveys. SkAI astrophysicists, foundational AI researchers, educators, AI ethicists, software engineers, and artists will work together to accomplish SkAI’s research goals and develop effective and transferable programs to grow a large and diverse Astro-AI and STEM workforce. SkAI will unite 25 partner organizations to form an inclusive, cross-disciplinary nexus centrally located in Chicago and the Midwest, with research and education bridges to Georgia, Hawaii, and Alaska, seeding and nurturing a diverse generation of interdisciplinary leaders in science and engineering. SkAI will leverage the fast-paced revolution in AI and the data revolution in astrophysics to revolutionize both fields.<br/><br/>SkAI researchers will tackle some of the grandest open questions in astrophysics, spanning 20 orders of magnitude in scales of time and space, probing the nature of stars, compact objects, and transients, galaxy formation and evolution, and the cosmic history of our Universe. The institute will achieve foundational AI breakthroughs in generative models, necessary for scaling to large, heterogeneous datasets and for combining simulations with data-based approaches to modeling; astrophysics-informed and interpretable architectures, needed to infer physically meaningful parameters and gain new insight into the information latent in large datasets; and uncertainty quantification, to guarantee reliable scientific results and conclusions. The SkAI team will integrate Astro-AI efforts and resources around three pillars that mirror the scientific-process elements of data, simulations, and experiments: enhanced inference from cosmic survey data, AI-accelerated simulations with multi-scale astrophysics, and learning-based astrophysical survey and instrument design. Through interactive workshops, summer schools, a vigorous visitors program at the SkAI Hub in Chicago, and the release of ethical and easy-to-use community software, SkAI will engender a broad collaborative basis for interdisciplinary science involving a diverse and inclusive group of AI researchers, astrophysicists, engineers, and students. This institute is jointly funded by the NSF and The Simons Foundation.<br/><br/>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.
