Energy eyes new data-curation org for science and engineering AI models
The Department of Energy is in the planning stages for a public-private consortium that will pull together scientific data across its national laboratories to train “self-improving” artificial intelligence models.
In a request for information published Thursday, the DOE said these models — to be used for science and engineering — will be made available through the consortium, via cloud technologies, to the U.S. government, academia and private-sector programs.
“This historic mobilization of DOE, the National Laboratories, and private partners will serve as a force multiplier in executing America’s AI Action Plan to achieve global dominance in AI, and to advance scientific discovery, energy, and national security,” the RFI stated.
Released in July, the Trump administration’s AI Action Plan has a robust energy-focused section that includes multiple callouts for DOE, the national labs and the country-wide buildout of AI data centers.
The plan specifically tasks DOE, the National Science Foundation, the National Institute of Standards and Technology and other federal partners to invest in “automated cloud-enabled labs” for various science and engineering fields. Federal agencies, the private sector and research institutions would be enlisted to collaborate and coordinate with the labs, while researchers would be incentivized to “release more high-quality datasets publicly.”
The RFI, meanwhile, seeks input from institutions on how best to develop “leading-edge AI models” that can leverage DOE data, facilities and expertise. The agency also puts a call out for interested partners, including think tanks, investors, research organizations, and AI developers that have developed AI models with “enhanced capabilities.”
The document poses a series of questions to interested parties, asking for recommendations on how DOE can “best mobilize” the labs to partner with industry and make sure the collected data is “structured, cleaned, and preprocessed in a way that makes it suitable for use in AI models.”
The RFI is also interested in how it should structure the public-private consortium in a way that ensures “activities across a range of scientific and technical disciplines” are properly enabled. And DOE would like to know how it can “best provide AI models to the scientific community through programs and infrastructure making use of cloud technologies to accelerate innovation in discovery science and engineering for new energy technologies.”
The new RFI follows another AI-related DOE request last month, which sought proposals for the buildout and maintenance of data centers and energy generation infrastructure on Oak Ridge National Laboratory’s campus in Tennessee.
