President Trump signed an executive order launching the Genesis Mission this week, which he's explicitly comparing to the Manhattan Project in terms of urgency. The stated goal: to use AI to double the productivity of American science within a decade.
This isn't just about curing cancer; it's about establishing government control over the infrastructure layer where AI training happens and maintaining technological dominance against competitors like China.
The American Science and Security Platform
The mission directs the Department of Energy (DOE) to build an American Science and Security Platform that integrates the nation’s 17 national laboratories' supercomputers and datasets into one unified AI experimentation system.
This platform is designed to let AI models design experiments, run simulations, and iterate autonomously, potentially shortening discovery timelines from years to days. The infrastructure already exists in labs like Oak Ridge and Los Alamos; Genesis is about building the integration layer that makes that compute and data seamlessly accessible to AI.
The Unglamorous Reality: Data Modernization
The actual, critical work of Genesis is incredibly unglamorous: data modernization. Decades of scientific data are currently sitting on tape systems—essentially behind a wall that AI can't access. Genesis is fixing this infrastructure gap so models can train on massive amounts of real, valuable experimental results, not just simulated data.
This kind of infrastructure investment—making federal data more accessible—is incredibly valuable and will pay dividends across scientific domains, even if it's less exciting than "AI discovers fusion energy".
The National Security Angle
The order explicitly ties Genesis to national security. The DOE oversees nuclear weapons through the National Nuclear Security Administration (NNSA), which is actively participating. The objective is to use AI, quantum computing, and advanced data analytics to strengthen deterrents and ensure an "unmatched strategic edge".
This is government infrastructure capture. By integrating federal resources, the DOE becomes the gatekeeper, locking in relationships with private partners like NVIDIA and OpenAI and defining the AI research landscape for years.
My Take
The Manhattan Project framing is hyperbolic, especially since the order doesn't specify a dedicated budget. However, the intent is sound: AI is now a strategic resource, and the government is stepping in to ensure it controls a major piece of the infrastructure required to advance it.
If Genesis successfully fixes the data infrastructure problem, it will be meaningful progress. The real measure of success won't be in marketing slogans but in whether it builds useful infrastructure that helps researchers work more efficiently. That's a low bar for a "Manhattan Project," but a high bar for a government IT program.