IA en un minutoNewsBusiness and products
Big tech companies will contribute $2.4 billion to the Genesis Mission, the US plan for AI-driven research
IA en un minuto newsroom · Editor: Jon Elgezabal

In 30 seconds
The White House has lined up eleven tech companies that will put $2.4 billion in computing and AI tools at the service of US public science. NVIDIA leads the list with $1 billion, followed by AMD, OpenAI, Anthropic and Google. Bloomberg notes it is still unclear how researchers will be guaranteed that computing time, and the government is adding a $100 million fellowship for fast-track PhDs.
The White House has announced that eleven technology companies have pledged a combined $2.4 billion to advance scientific research through artificial intelligence.
The Trump administration unveiled the pledges on Thursday at a science summit held in Washington. They consist of AI tools for science (superintelligence, in the White House's wording) and computing credits for the Genesis Mission Consortium, the US government's plan for AI-powered research, and they will support more than 15 federal agencies. The largest contribution, $1 billion, comes from NVIDIA; AMD is putting in $500 million and OpenAI $200 million, while Anthropic and Google will give $150 million apiece. AMP and Emerald AI are each contributing $100 million, and Armada, AWS, Crusoe and Micron $50 million each.
The White House says the pledges come after the mission was widened in July to more than 15 federal agencies, and after a federal commitment of $5 billion to national science and technology challenges in energy, health, space and other areas. Politico, in an article carried by Yahoo that cites statements from the administration and the companies, gives some detail: NVIDIA's part is computing capacity over the coming five years; OpenAI's is discounted tokens and training programs for scientists; Anthropic's is three years of access to Claude; and that of AWS is cloud credits.
At the same summit the government launched the Genesis Mission Fellowship. The $100 million program will let students complete accelerated PhDs in four years with a dual grounding in superintelligence and in a second scientific or engineering discipline, including research placements in industry and at one of the Department of Energy's national laboratories.
Bloomberg reports that it is unclear how that computing capacity will be reserved for scientists, given that the major cloud providers already face strong demand and lucrative prices for computing power.
Why it matters · analysis and opinion
That the same chipmakers and AI labs competing to sell computing are contributing it to public science says a lot about how power in research is being shared today: whoever controls the data centers sets the pace. For the companies, it is also a way to place their models and chips inside federal agencies and to train scientists on their tools. What is missing for it to have a real effect is the how, which part arrives as guaranteed capacity and which part as discounts or credits that depend on free room in data centers already in high demand. And it confirms that the race for scientific AI is increasingly played out through deals between the government and big tech, not only with public budgets.
Official source: White House · Written with the help of AI: how we make the news


