As the worldwide AI gold rush continues apace, the United States government continues to signal an urgent need for American leadership in this still-murky field. Yesterday, President Trump signed a new executive order outlining what the administration calls the Genesis Mission, an initiative that seeks to spur a nationwide effort to make the U.S. the world leader in AI technology and its practical application.
The White House's press release emphasizes the importance of its existing AI Action Plan, and compares the scale of its effort to the Manhattan Project, the initiative that produced the atomic bomb during World War II.
According to the new executive order, the plan is to be carried out by the Department of Energy (DoE), which controls key supercomputing resources in the USA among its other responsibilities. Secretary Chris Wright states: "The Genesis Mission will dramatically accelerate scientific discovery, strengthen national security, secure energy dominance, enhance workforce productivity, and multiply the return on taxpayer investment into research and development." There's no word on how the Genesis Mission is meant to be funded.
In its ideal form, the project is tasked with building an AI platform to collect federal scientific datasets to train AI models for the practical goal of solving at least 20 technological challenges. These will be collected among the areas of materials, manufacturing, biotechnology, nuclear fusion and fission energy, quantum information, semiconductors, and electronics.
Given the Genesis Mission is meant to be coordinated among several agencies and private partners, additional challenges can be proposed by those stakeholders. The challenges will be reviewed and adjusted yearly.
The first stage of the project will be to identify logical and physical resources, namely computing infrastructure, datasets, and models. The Genesis Mission could to make use of the supercomputers at DoE's national laboratories, but will also collaborate with agencies and partners "possessing advanced AI, data, or computing capabilities or scientific domain expertise."
That's a slightly curious statement, as most any capacity for datacenter hardware and AI chip manufacturing is de facto signed for for multiple years, meaning it's unclear whether the DoE wants to build out its own datacenters, rent existing capacity, or, most likely, both. An unnamed source at the White House told the New York Times that Nvidia, AMD, HP, and Dell have agreed to build facilities within national laboratories.
Stay On the Cutting Edge: Get the Tom's Hardware Newsletter Get Tom's Hardware's best news and in-depth reviews, straight to your inbox. Contact me with news and offers from other Future brands Receive email from us on behalf of our trusted partners or sponsors
As for model training data, the project wants to ensure access to "appropriate datasets, including proprietary, federally curated, and open scientific datasets, in addition to synthetic data generated through DOE computing resources". Interestingly, the document makes repeated mentions of intellectual property protection and provenance tracking, presumably in the context of public-private partnerships, including "innovations arising from AI-directed experiments."
A program of this scale adds to the growing concerns about dwindling supplies and high prices for electricity, prompting the DoE Secretary to state the project should “make [the U.S.'] electricity grid more efficient and reverse price rises that have infuriated American citizens." The matching press release at energy.gov clearly states the Mission "will accelerate advanced nuclear, fusion, and grid modernization using AI to provide affordable, reliable, and secure energy for Americans."
... continue reading