The National Working Group on Artificial Intelligence Research released a new report Wednesday outlining a plan to develop a standardized cyber infrastructure for artificial intelligence research, hoping to promote responsible AI applications.
Fair access to data from computational research is a key feature of the group’s report. NAIRR co-chairs discussed how researchers associated with rich private companies or universities with stable donations have disproportionate access to computing power and data for advanced research.
“We need to expand the ability of the American research community to pursue cutting-edge research on artificial intelligence and make the paths fairer,” said Lynn Parker, co-chair of NAIRR, during a news conference. “NAIRR seeks to reduce barriers to entry to allow more Americans to engage in AI research and development to prevent unintended consequences or different impacts of AI use.
IN interim report–– the first of a series of two reports–– consists of recommendations for federal agencies to refer to when studying new artificial intelligence technologies. Using NAIRR’s own user access portals, researchers will be able to access a diverse amount of computing resources, as well as connect users to the cloud and high-performance computing servers.
Manish Parashar, a fellow NAIRR co-chair, told reporters that user portals would also contain datasets to give researchers access to a “stable, transparent and responsible data ecosystem”.
“NAIRR will also act as a hub for AI test benches and test datasets, cataloging and providing access to existing resources for testing and benchmarking that can accelerate progress in AI R&D [research and development] and increase the research community’s access to these high-quality test beds, “he said.
Given the large amount of potentially sensitive data stored in the new network, the working group also advocated the use of a standardized zero-confidence architecture to protect databases.
While NAIRR management wants to create incubator resources to accelerate the development of new artificial intelligence technologies, the report also highlights best practices for responsible and ethical AI.
“As a nation, we must unite to expand access to the resources that fuel AI, providing ways for more Americans to engage in research and development (R&D) and access to state-of-the-art resources,” the summary said. of the report. .
These ways to eliminate threats to civil liberties through artificial intelligence technologies include better testing procedures, mitigation of prejudices and improved security features. Incorporating different perspectives and experiences into the development of artificial intelligence applications is also a priority in the NAIRR report. Parker noted that maintaining accessible datasets and computing power will help make AI development widely available.
“The strategic goal we set at NAIRR would be to strengthen and democratize AI’s innovative ecosystem in the United States in a way that protects privacy, civil rights, and civil liberties,” she said. “AI is transforming our world, and the growing division of resources between those who have access to the resources needed to pursue cutting-edge AI and those who do not threatens our nation’s ability to cultivate a research community and a workforce that reflects a rich Diversity of America ”