DRIVING CONSTANT INNOVATION

The foundation of this portfolio should be a metascience unit in each federal science agency. Each unit should be highly empowered, reporting directly to the director or administrator to ensure cross-agency visibility and guard against capture by particular programs or constituencies. Each unit should be staffed with researchers possessing expertise in the science of science, program evaluation, and data analysis, supplemented by rotating program officers who bring operational knowledge of how grants actually get made. Federal funding agencies should develop systematic gap-mapping capacity, regularly review their funding portfolios, and drive more intentional grantmaking instead of deferring to the portfolio allocation of the previous fiscal year. Such a process could identify both bottlenecks and the foundational capabilities that would address them.80 Metascience units could, for instance, convene expert workshops and maintain living maps of capability gaps, or work externally with foundations that have developed sophisticated methods for identifying transformative research opportunities. DARPA’s Heilmeier Catechism embodies this discipline, forcing explicit articulation of what gap a program addresses and why solving it matters.81

These units should also be empowered to do more than advise on the existing portfolio of instruments; they should pilot new ones across agency programs.

An NIH unit might randomize whether study sections use golden tickets, then track the novelty and citation impact of funded projects across conditions. An NSF unit might compare outcomes from fast grants against standard review timelines. A DOW unit might experiment with how much discretion program managers are given in funding decisions, comparing data across branches. Without the authority to run experiments and to compel program offices to participate, these units will devolve into compliance operations producing reports. One way to secure this authority is to give each metascience unit a budget it can regrant to program managers for participating in experiments. Their findings should also be published externally, building the broader evidence base on what works in science funding and creating accountability to act on what is learned. The United Kingdom’s Metascience Unit, established in 2024, offers an early model, reporting in its first year on distributed peer review, partial randomization of awards, and the consistency of reviewer judgments.82 All this institutional experimentation must be matched by hiring the highest quality staff.

We can, and must, make program management one of the most sought-after jobs in science, where talented people can shape the direction of entire fields. DARPA’s success rests not on any single mechanism but on hiring the right program managers and giving them genuine discretion. This begins with making it easier for people from a wide range of backgrounds, including industry and philanthropy, to enter a short stint in government, and with raising the prestige and profile of program officers, whose efforts in coordinating entire fields toward major breakthroughs often go underrecognized. We need to recruit the best scientific talent into these roles, and then truly empower them, with resources, freedom, and the opportunity to network with the smartest people tackling the hardest problems. America invented the modern research architecture with institutional innovations that the rest of the world subsequently adopted. We must lead the charge again.

WE MUST CHOOSE OUR TECHNOLOGICAL FUTURE

In centuries past, land and population determined national power. The industrial age added capital and manufacturing capacity as fundamental components of national sovereignty and security. As the President’s National Security Strategy makes clear, technological capability has always conferred advantages, and now sets the terms on which all of these inputs operate.83 Technological leadership helped forge modern America. Science alone did not produce this leadership; it required the deliberate cultivation of engineering talent, institutional capacity, and industrial might to turn discoveries into capabilities. Our physicists translated Schrödinger’s equations into the weapons that ended World War II and defined the nuclear order that followed. Our engineers turned Shannon’s information theory into the protocols that carry the world’s digital communications. Our scientists turned advances in physics and materials science into GPS satellites that guide ships, planes, and precision weapons on every continent. Technology has become foundational to a nation’s economic and military strength, its capacity to act independently in the world, and its ability to maintain its distinct culture and way of life. American technological leadership produced enormous wealth, secured our homeland, and turned our nation into a beacon for the rest of the world. It cannot be taken for granted. The nature of technological advantage is shifting. As will be discussed in Chapter V, advances in AI expand the world’s ability to generate ideas and will accelerate scientific research. These capabilities will benefit American scientists. But those benefits will also accrue to our competitors. As new tools of discovery become more widely available, the comparative advantage conferred by scientific excellence alone will likely narrow. It will therefore be equally important for our nation to bolster its capabilities in translation, the process turning ideas into realworld capabilities. That is the subject of this chapter.

FAILURE OF THE PASSIVE MODEL

America has long been the world’s most prolific source of scientific breakthroughs. We must ensure we are equally formidable at turning those breakthroughs into national power, or we risk watching the fruits of American discovery harvested first by others.

For decades, American science and technology policy rested on an unspoken assumption that government need only fund basic research, support a vibrant economy, and trust that technological strength would follow. Pour money into universities, protect intellectual property, keep markets open, and the innovations that secure the nation and enrich its people arrive on schedule—this was the implicit bargain of the post-war scientific order, and for a generation it appeared to work. That laissez-faire assumption does not survive contact with competitors who have built technological states. Commerce and research now constitute a geopolitical battlespace, and the parallel to trade policy is instructive.

For decades, the United States assumed that open markets would naturally produce American prosperity and that free trade would lift all boats by maximizing global efficiency. Instead, unilateral openness hollowed out the American industrial base. Competitors exploited our markets while protecting their own. The gains from trade accrued to a narrow slice of the economy while entire communities lost their livelihoods. President Trump has delivered a necessary correction, recognizing that economic security is national security. The same logic applies to science and technology. The assumption that federal research investment alone would sustain American technological dominance has proven naive. We funded the discoveries, trained the researchers, and published the papers, but did not ensure that the benefits accrued to our nation. The ideas, as well as the time- and resource-intensive parts of the development cycle, are taken abroad to benefit others. American researchers invented the flat-panel display; Asian manufacturers captured the market. American scientists pioneered cutting-edge battery chemistries; production scaled overseas. The pattern has repeated across decades and industries. The cause was the same passive model that hollowed out our factories, while our competitors pursued a holistic strategy that deliberately blurred the line between public and private, civilian and military, treating every advance, wherever it originated, as raw material for their state-directed development.

CHOOSING TO LEAD

Our nation’s technological outcomes are shaped by policy choices. The internet became an American platform because we embedded openness and competition into its foundations. We chose to go to the Moon in 1969 because of our national will. It is an achievement that appears, in retrospect, jarringly out of place in humanity’s technological timeline. Conversely, nuclear energy stalled in America not because the physics failed, but because regulatory choices over the past half-century made building uneconomical. In each case, the decisive variable was the set of institutional, regulatory, and strategic choices that determined whether science became capability. Competition may dictate that nations will adopt AI, race from genotype to phenotype, pursue nuclear technology, and build advanced warships, but it does not dictate how they go about it, or even necessarily when. Within broad technological trajectories, multiple futures are possible. The question is which one America will fight for.84

FIGHTING IN OUR OWN ARENA

No nation, however powerful, can lead in every domain. Some technologies demand that we press forward, extending strengths into durable advantages where early leads compound over time. Others require that we hold ground. We may not seek total dominance, but we will not permit an adversary to achieve it either. In domains of lesser strategic consequence, we can concentrate our energies elsewhere and ensure that strengths accrue to partners, rather than adversaries.

The technologies that matter most are those that form platforms on which future technologies are built. Dominating the right foundational platforms grants structural power, allowing the leading actor to dictate the rules and standards by which others must play. These advantages compound, with advances in one field, like computation, unlocking breakthroughs in others, like AI and biotechnology, creating feedback loops that reinforce the leader’s edge.85 The semiconductor industry offers one striking example. It was not predetermined by nature that transistor density would increase exponentially over time, as described by Moore’s Law. Rather, the market found ways to harness, fund, and develop the miracles of physics and material science that made such exponential growth possible. In turn, the countries that control the semiconductor supply chain now push a snowball that gets bigger every year, incorporating more and more of the modern economy. Mobile communications, satellite navigation, and ever-more powerful AI systems are all built on this platform.

Winning does not mean isolation. Yes, sovereignty over critical technologies often means building the capacity to produce them domestically. But as the semiconductor example illustrates, technological leadership can also mean strategic integration, supplying the platforms that drive economic growth across our partners’ economies, like tapping into broader markets to push the snowball of Moore’s Law. Leadership means charting our own destiny by choosing which technologies we develop, which standards we set, and which supply chains we control, rather than accepting a world shaped by the choices of others. Accomplishing those goals requires that we understand and amplify our true advantages.

Among our competitors, some, despite their industrial strength, lavishly fund state priorities while leaving private markets to languish, missing breakthrough technologies that emerge from the fringes. Their economies remain investment-driven rather than productivity-driven, with total factor productivity contributions declining even as R&D spending rises. And while the success of their consumer applications is often mistaken for genuine R&D-intensive innovation, they remain far behind the United States in driving original frontier research.86 Meanwhile, America’s financial architecture channels capital toward frontier technology at a scale no nation can rival. In 2024, American venture capital firms deployed over $200 billion, accounting for 57% of global venture investment.87 Our public markets tell an even more striking story. As the time of writing, the seven largest American technology companies are collectively worth more than the entire stock market of our primary competitor. These figures reflect not just deep pools of capital, but liquid markets that reward successful exits, a legal system that enforces contracts and protects property rights, institutional investors with long time horizons, and a startup ecosystem that treats failure as education rather than disgrace. This is our arena. The task before us is to match the incredible vibrancy of our markets with our scientific capital, to unleash technological capabilities that benefit the American people.

UNLEASHING INNOVATION

Over the past few decades, America has built a regulatory state that brings down a gavel to block much innovation in the physical world. When it takes longer to obtain a permit than to build the thing being permitted, when the default answer from the government is “no” or “wait,” the most talented builders go elsewhere or stop trying. This system selects against the kind of people and organizations that drive innovation, namely small teams, unconventional entrants, and entrepreneurs whose opportunity costs are so high that they will not wait years for approval.

The nuclear industry illustrates the problem at its most extreme. The United States once led the world in nuclear technology. We built the reactors, trained the engineers, and wrote the safety standards that other nations adopted. Then we regulated the industry into paralysis. Before an advanced reactor startup can pour a single foundation, it must spend five to six years in pre-application discussions with the Nuclear Regulatory Commission (NRC), followed by a design certification process that can take an additional four years. One company had to submit a 12,000-page application, supported by more than 2,000,000 pages of technical documentation. DOE spent over $600 million in funding to support this process for a single reactor design. Then came the combined license application, with its own multi-year safety and environmental reviews, mandatory public hearings, and construction inspections.88

THE FREEDOM TO BUILD

The first Trump Administration began to fix this regulatory morass in the nuclear realm, supporting bipartisan legislation to reform the NRC’s approach.89 In the second Administration, we have moved to break the logjam entirely because the stakes are so high, driven by energy demands across America’s AI and manufacturing industries.

In May 2025, the President signed four executive orders overhauling America’s nuclear regulatory framework: imposing an 18-month deadline for the NRC to revise its regulations, capping licensing timelines for new construction applications, creating expedited approval pathways for reactors already tested by DOE or DOW, and directing the NRC to weigh the benefits of nuclear energy to economic and national security in its regulatory decisions. Companion orders direct DOE to facilitate five gigawatts of power uprates to existing reactors, begin construction on ten new large reactors by 2030, and invoke Defense Production Act authority to secure domestic nuclear fuel supply chains. These are the most sweeping nuclear reforms in a generation, taking down the old system that placed the status quo above the American people.90 In biotechnology, the United States pioneered many of the foundational advances in genomics, gene therapy, and CRISPR-based medicine, yet our clinical trial system has grown so costly that testing American discoveries increasingly happens abroad. A promising retinal prosthesis that restores vision for the blind, developed in Alameda, California, had to run its clinical trials in Europe due to challenges navigating the approval process.91 Per-patient costs for clinical trials run far higher than in other economies; American scientists make the breakthroughs, but the infrastructure to validate and deploy them is migrating overseas.92

Here, we have also begun to reverse the trend. The U.S. Department of Health and Human Services (HHS) has launched the largest deregulatory effort in the Department’s history. The Food and Drug Administration (FDA) has moved to accept real-world evidence in regulatory reviews,93 dropped the default requirement of two clinical trials per drug application in favor of a single well-powered study with confirmatory evidence,94 and fast-tracked review timelines for drugs supporting U.S. national interests. NIH introduced a new site to make the community aware of priority scientific areas without the need for new Notices of Funding opportunities and also eliminated application requirements that added burden without commensurate benefit.95 These reforms are essential, and they must mark the start of a sustained effort to ensure that the world’s most innovative biomedical science is tested and deployed on American soil.96 These examples should only be the beginning. The best of American innovation has always been characterized by permissionless experimentation, the freedom to build, test, fail, and try again without asking leave at every step.

A permissionless approach to innovation does not mean the reckless development of technology. Prudence in broad deployment is wise, and it is the foundation of society’s trust in our technologies. But policymakers must also price in the harms of stagnation: the economic growth foregone, the lives lost waiting for a cure, the industrial and automotive accidents that happen by failing to adopt more advanced technology. Where existing rules do not fit new technologies, regulatory sandboxes that allow real-world testing under controlled conditions can generate the evidence needed to write sensible ones. Our goal should be to dismantle the procedural obstacles that prevent American knowledge from becoming American technology, while maintaining genuine accountability for results.

PLACES TO TEST

About seventy years ago, a committed group of amateur rocketeers purchased a private test site in the Mojave Desert north of Edwards Air Force Base. Since then, the oldest continuously operating amateur rocket group in the country has been firing homemade engines. In 2003, a spin-off organization incorporated 39 Chapter III – Securing U.S. Dominance in Critical and Emerging Technologies next door as a nonprofit, and its volunteers built the necessary infrastructure from scratch. They erected static test stands, reinforced concrete blockhouses, and propellant storage sites. They secured federal permits for handling high explosives and a Federal Aviation Administration (FAA) waiver to launch rockets to 40,000 feet on weekends. A reservation there costs anywhere from a few hundred dollars to the low thousands.97 On any given Saturday, at this facility, a father-and-daughter team working out of their garage may be standing next to university engineering students and off-duty aerospace professionals, all firing liquid engines in the open desert. The adjacent airport itself became an FAA-licensed commercial spaceport in 2004.98 Multiple rocket companies have emerged from this cluster of cheap leases, federal licenses, and shared test infrastructure to win NASA prizes, raise substantial venture capital, and reshape the commercial launch industry.99 Mentors at the amateur site were recruited into startups. Startup veterans returned to mentor the next cohort of rocket enthusiasts. These organizations have overcome the odds in one of the most heavily regulated industries in America and have bred a vibrant ecosystem for innovation.

Today, the Federal Government also offers test stands at industrial scale through the Stennis Space Center, where startups can lease facilities rather than spending tens of millions of dollars building their own.100 Providing test infrastructure and sustaining a regulatory environment that allows our innovators to experiment represents one of the most important levers we have for driving technology forward.

Test infrastructure, whether for rockets, advanced manufacturing, autonomous systems, or any frontier technology, is an enabling resource that determines whether the next great American company starts in a desert lot or dies on the vine in a student’s garage. Wherever AI systems are deployed at scale, wherever advanced reactors are built and operated, wherever synthetic biology is used in agriculture and medicine, the resulting standards, supply chains, and knowledge bases will compound in favor of the nation that moved first. America must be the place where that experimentation happens.

While bright spots like the Mojave site and the Stennis Space Center exist for particular industries, across much of America, the gauntlet between a scientific discovery and a deployed technology has grown so forbidding that many of our best ideas never make it through. Too often, ideas die after the paper is published. A breakthrough in a university laboratory must be prototyped, tested under real-world conditions, validated against safety and performance standards, manufactured at scale, and brought to market. The most talented builders increasingly migrate toward software, where the regulatory burden is lightest. This contributes to the lopsided economic growth we see today, away from the world of atoms in our heartlands, and toward the world of bits in the Valley.