US science gets a new blueprint 81 years after 'Endless Frontier'

Eighty-one years ago, Vannevar Bush handed then-US President Harry S. Truman a report that would define American science for generations. "Science: The Endless Frontier" made the case for federal funding of basic research – and delivered the institutional architecture that helped win World War II, put a man on the Moon and gave the world the digital age. Now the White House has released its successor.

"Science: A New Golden Age," submitted to President Donald Trump on July 21 by Michael Kratsios, director of the White House Office of Science and Technology Policy, is the most sweeping reset of US science policy in eight decades. It argues that the postwar model – the "linear model" of basic research flowing into applied work and then into industry – no longer fits reality. The new report says discovery today is an iterative loop, and the institutions built for mid-century science are increasingly a bottleneck.

From peer review to 'golden tickets'

The report's first major thrust is institutional reform. It calls out a system where researchers spend nearly half their time on paperwork, where grants take the better part of a year from submission to award, and where peer-review panels tend to reward consensus over bold ideas.

Its prescription: put the individual scientist back at the center. Expand portable graduate fellowships and long-horizon grants modeled on the NIH director's Pioneer Award. Diversify funding mechanisms beyond peer review – including "golden tickets" that let individual reviewers champion ambitious proposals, fast grants for rapid decisions and prize challenges that pay for results. The report also calls for the creation of new institutional forms: X-Labs, agile time-bound teams for specific bottlenecks, and Advanced Research Projects Agencies (ARPAs) where program managers make bold bets.

The message is clear – the US government spends about $200 billion annually on R&D, and the report's authors believe too much of it is going to bureaucracy rather than breakthroughs.

The Genesis Mission: AI as a scientific engine

The most futuristic chapter is reserved for what the report calls a "revolution in science" powered by artificial intelligence (AI). Its flagship is the Genesis Mission, an AI-for-science initiative that aims to double US research productivity and impact within a decade.

Genesis would integrate supercomputers, AI models, scientific instruments and datasets across the national laboratory system – creating a closed loop where AI generates hypotheses, automated laboratories test them and the results feed back into the model. The report calls for building verification infrastructure at scale, noting that "while the cost of generation has decreased exponentially, the cost of verification has not." It also backs autonomous experimentation – self-driving labs that can collapse discovery timelines by orders of magnitude.

The ambition mirrors what China has been building in AI-driven science, from Baidu's PaddlePaddle in computational biology to Huawei's Pangu weather models. But the US version comes with a policy twist: the report simultaneously mandates "Gold Standard Science" through executive order, enforcing reproducibility and data sharing as the foundation for AI-powered discovery.

The competition that isn't named

China is rarely mentioned by name in the report, but its presence looms over every chapter.

"Our competitors are channeling unprecedented resources into science and engineering, taking a whole-of-society approach to seize the high ground in strategic technologies," the report says. It argues that scientific leadership alone no longer guarantees economic or national security – breakthroughs must be prototyped, manufactured and scaled domestically. This is the logic behind the report's push to strengthen public-private partnerships, open federal labs to industry and build dense regional innovation clusters tied to advanced manufacturing.

The strategic shift is notable. While the first Trump administration focused on tariffs and trade wars, and the Biden administration on export controls and the CHIPS Act, this report represents something different – a systematic attempt to rebuild the entire US innovation ecosystem from the ground up. The emphasis is less on blocking competitors and more on removing the bottlenecks that have slowed American science itself.

Can it deliver?

The report's ambition is undeniable, but its path is uncertain. Many of its recommendations – cutting bureaucracy, streamlining peer review and opening federal facilities – require legislative action or deep institutional change that has resisted reform for decades. And the $200 billion federal R&D portfolio is spread across agencies with entrenched cultures and constituencies.

What is clear is the direction: after 81 years, the US is declaring its postwar science model obsolete and betting on a new one built around AI, individual talent and industrial integration. For the rest of the world – including China, which the report implicitly treats as the benchmark for competition – the question is not just whether America can rebuild its own engine, but what the new rules of the global science race will look like when it does.