The National Science Foundation (NSF) disclosed details today regarding the pilot initiative for the National Artificial Intelligence Research Resource. This program involves a collaboration between the NSF, other federal agencies, and the private sector to provide researchers with access to datasets, AI models, and training opportunities.
Sethuraman Panchanathan, the Director of the NSF, highlighted that the NAIRR pilot program, introduced in November last year, aims to democratize access to generative AI (genAI) for organizations that do not have the financial resources of major tech companies or government entities. He emphasized during a Zoom briefing, “This is how we ensure equal opportunities, foster thriving industries of the future, and position ourselves at the forefront of competitiveness, societal progress, and prosperity on a global scale.”
The primary objective, as outlined by Katie Antypas, the chairwoman of the NSF Office of Advanced Cyberinfrastructure, is to facilitate cutting-edge exploration into fundamental genAI concepts and the application of AI in operational and web sciences. She expressed the intention to expand participation in the AI landscape to previously underserved domains.
To accomplish this, the NAIRR pilot program will grant US academics and researchers access to state-of-the-art technologies, AI datasets, models, and training resources. For instance, domain experts in fields like meteorology could leverage datasets from NOAA and NASA for AI analysis, while educators in underserved areas could utilize AI-centric educational tools such as digital notebooks and remote computing systems to enhance student learning experiences.
Antypas noted that the NAIRR initiative aims to address the disparities in genAI adoption across different communities in the US. She highlighted the challenges faced by individuals and groups lacking significant resources or connections to access the necessary data and tools for effective utilization of genAI technologies.
The NSF’s pilot program is scheduled to span two years and will encompass four key components: NAIRR Open for open AI research utilizing diverse resources, NIR Secure focusing on privacy and security aspects of AI development, tools and platforms research, and NAIRR Classroom tailored for educational purposes.
Researchers will be able to access NAIRR’s resources through an online portal, with a broader call for proposals expected in the spring of 2024.
