The Great Reinvention logo THE GREAT
REINVENTION

The American Mind Project: A National Research Network for AI, Biotech, and Human Flourishing

Proposed legislation: The American Discovery and Human Flourishing Act

The American Mind Project: A Cost-Benefit and Investment Analysis

For three-quarters of a century, the engine of American prosperity has been discovery. The transistor, the internet, GPS, the human genome, GPS-guided agriculture, mRNA vaccines, and the foundational mathematics behind today's artificial intelligence all trace back, in whole or in part, to publicly funded research. That research did not merely produce knowledge; it produced industries, jobs, and decades of economic growth. Yet today the United States is allowing its lead to erode at precisely the moment when the next general-purpose technologies — artificial intelligence and biotechnology — are arriving.

The American Mind Project is a proposal to renew that engine deliberately and at scale: a national research network connecting universities, national labs, and public-private institutes focused on artificial intelligence, biotechnology, and the broader project of human flourishing — linked physically by high-speed rail between research hubs, and socially by a large national program of research fellowships. Funded at roughly $15–30 billion per year, it would treat scientific and human-capital investment as the strategic infrastructure it has always been.

This page examines the case honestly: why the investment is needed now, what it would build, what the historical returns to public research actually are, and where the genuine risks and counterarguments lie.

The Need: A Slipping Lead at the Worst Possible Moment

The headline numbers are sobering. According to the National Science Foundation's National Center for Science and Engineering Statistics, the United States invested about 3.5 percent of GDP in R&D in 2021 — strong, but behind leaders like Israel and South Korea (both above 4 percent). More important than the level is the trend: NSF data show that China's R&D spending grew nearly twice as fast as America's from 2019 to 2023 (roughly 8.9 percent average annual growth for China versus 4.7 percent for the U.S.), and by some measures, as the Information Technology and Innovation Foundation (ITIF) has reported, China has caught up to or surpassed the U.S. in gross R&D and is closing fast in advanced-industry research.

There is also a composition problem. In 2023 the federal government funded only about 18 percent of total U.S. R&D, with business funding roughly 75 percent. Business R&D is overwhelmingly applied and near-term; the high-risk, long-horizon basic research that seeds entirely new industries is exactly what the private sector underinvests in — and exactly what public funding has historically supplied. Even within the new programs Congress created to respond, the follow-through has faltered: lawmakers authorized $20 billion over five years for NSF's Technology, Innovation and Partnerships (TIP) directorate but, per ITIF, appropriated only about $410 million of it.

The American Mind Project is designed to fill precisely this gap: sustained, mission-driven, public investment in the basic and translational research that private markets won't fund but that the nation's long-term competitiveness, security, and well-being depend on.

What Gets Built

The project has four interlocking components.

1. A national network of research hubs. Rather than concentrating funding in a handful of coastal superstar institutions, the program creates and strengthens research centers across the country — at universities, national laboratories, and new public-private institutes — specializing in AI, biotechnology, and adjacent fields. The explicit goal is geographic breadth, spreading both opportunity and resilience.

2. Public-private mission labs on the DARPA model. The most successful innovation institution in American history is arguably DARPA, the Defense Advanced Research Projects Agency. Created in 1958 to prevent "technological surprise," DARPA pioneered or seeded the internet, GPS, stealth, drones, voice interfaces, and early mRNA vaccine work — what The Economist called "the agency that shaped the modern world." Its formula is well studied: empowered program managers, small teams, flexible contracting, aggressive timelines, and tolerance for high-risk failure. The American Mind Project would establish DARPA-style mission labs for AI safety and capability, biotechnology, and human flourishing, applying that proven model to civilian grand challenges.

3. A national fellowship program. Talent, not money, is the binding constraint on frontier research. The project funds a large cohort of research fellowships — graduate, postdoctoral, and mid-career — to train and retain scientists and engineers, and to keep the world's best minds working in the United States rather than abroad.

4. High-speed rail linking research corridors. To knit the network together, the program invests in high-speed rail between major research hubs, enabling the dense, face-to-face collaboration that drives discovery and binding distant institutions into a single innovation ecosystem. (As the cost analysis below makes clear, this is the component requiring the most caution.)

Cost Breakdown

At $15–30 billion per year, the program is large but proportionate — roughly the scale of a single major federal science agency's budget, and a fraction of the R&D gap relative to GDP. A representative allocation:

The Benefits: The Documented Returns to Public Research

The case for research funding rests on an unusually strong empirical record.

The Human Genome Project is the canonical example. The federal government invested about $3.8 billion through the project's completion in 2003. Economic analyses (cited by the National Human Genome Research Institute and others) estimated a return on the order of 141 to 1 — every federal dollar generating roughly $141 in economic activity — and broader accounting attributed nearly $1 trillion in economic growth and hundreds of thousands of jobs to genomics. Even the more conservative ongoing estimates put the field's return at roughly $4.75 for every $1 spent annually. (It is worth noting, in the spirit of honesty, that FactCheck.org and others have cautioned that the headline ROI figures rest on input-output multipliers that some economists consider generous; the directional conclusion — large positive returns — is robust, even if the precise multiple is debatable.)

DARPA's portfolio seeded technologies — the internet and GPS chief among them — whose cumulative economic value is effectively incalculable and certainly in the trillions. No private firm would or could have funded the internet's foundational research on a commercial timeline.

Broader economic research consistently finds high social returns to public R&D. The Center for American Progress, ITIF, and academic studies generally estimate social rates of return on public research well above those of most other public investments, with ITIF specifically modeling that cuts to federal R&D measurably reduce long-run GDP growth.

Strategic and security benefits. Leadership in AI and biotechnology is not merely an economic question but a national-security one. Ceding the frontier in these dual-use technologies to a strategic competitor carries risks that no spreadsheet fully captures.

Human flourishing. Beyond GDP, the research targeted here — disease cures, mental-health treatments, clean energy, AI tools for education and medicine — bears directly on length and quality of life, the ultimate return on investment.

Administrative and Implementation Considerations

Money alone does not produce discovery; institutions do. Key design choices:

International Comparisons and Precedent

The model has strong international and historical validation.

The U.S.'s own postwar record — Vannevar Bush's "Endless Frontier" vision, the founding of NSF and NIH, DARPA, and the Apollo program — is the original proof that sustained public research investment can generate generational prosperity and technological dominance.

China is the contemporary cautionary tale: as NSF and ITIF data document, it has scaled R&D aggressively and is now at or near parity in gross spending, with explicit national strategies in AI and biotech. The American Mind Project is, in part, a response to that competitive reality.

Smaller nations show the model's reach: Israel and South Korea, with R&D intensities above 4 percent of GDP, have built outsized innovation economies. The European Union's framework programs and the proliferation of "ARPA-style" agencies worldwide (the UK's ARIA, among others) reflect a global consensus that the DARPA approach works.

The honest lesson is twofold: sustained investment matters, but institutional design matters as much as dollars — which is why the project emphasizes the DARPA model and stable funding, not merely a larger budget.

Comparison to the Status Quo and Alternatives

The status quo is gradual relative decline: a federal R&D share that has drifted down as a fraction of GDP, authorized programs left unfunded, and a growing reliance on near-term private R&D that neglects foundational science. It is a slow erosion whose costs appear only years later, as industries and talent migrate elsewhere.

Alternatives include:

The American Mind Project is best understood as restoring the basic-and-translational layer of the innovation stack — the layer that markets won't supply and that everything downstream depends on.

Risks, Trade-offs, and Counterarguments

A credible case must engage the strongest objections.

Research ROI figures are uncertain and sometimes inflated. The most important caveat: the eye-popping returns attributed to the Human Genome Project rest on economic-multiplier methods that, as FactCheck.org and some economists note, can overstate causation and double-count. The honest claim is that public research delivers large, positive returns on average — not that any specific program will return 141 to 1. Some bets will fail; that is inherent to frontier research and is a feature, not a bug, of the DARPA model, but it means the program must be judged on its portfolio, not individual projects.

High-speed rail is the weak link. This is the program's most vulnerable component on cost-benefit grounds. The Cato Institute and others argue persuasively that U.S. high-speed rail is so much more expensive than international peers — and so slow to build — that it risks being a money sink. The defensible position is to treat rail as a small, disciplined, reform-conditioned element, and to be willing to scale it back sharply if costs cannot be controlled. Tying a strong research program to a troubled rail program is a real political and fiscal risk.

Money can be spent poorly. Research funding can ossify into bureaucracy, reward incumbents, and chase fads. Without the DARPA model's autonomy and rigorous peer review elsewhere, the program could underperform its potential. Institutional design is decisive.

AI and biotech carry serious dual-use dangers. Accelerating these fields without commensurate investment in safety and biosecurity could increase catastrophic risk. The "human flourishing" mission obligates the program to fund guardrails as seriously as it funds capability — and critics will rightly judge it on whether it does.

Returns are long-delayed. The payoffs from basic research arrive over decades, making the program politically vulnerable to short-term budget pressure — the very dynamic that gutted the TIP directorate. Sustaining commitment across administrations is a genuine challenge, not a solved problem.

Talent retention depends on factors outside the program. Fellowships are wasted if visa and immigration policy drives trained researchers abroad. The program's success is partly hostage to policies it does not control.

None of these is disqualifying for the research core, but they argue for disciplined design: protect the DARPA model, keep peer review, fund safety alongside capability, stabilize the money, and treat the high-speed-rail component with particular caution.

Conclusion

Nearly everything that made the modern American economy — the internet, GPS, the genome, the foundations of AI — came from public investment in discovery. That engine is now running below the level the moment demands, even as China scales its own research nearly twice as fast and the next general-purpose technologies arrive. The American Mind Project, at $15–30 billion per year, proposes to restart the engine deliberately: a national network of research hubs, DARPA-style mission labs in AI and biotech, a large fellowship program to grow and keep talent, and rail links to bind the system together.

The historical returns to this kind of investment are, by any honest accounting, large — even if the precise multiples are debatable and individual bets will fail. The strongest version of the program protects the institutional features that actually generate discovery (autonomy, peer review, stable funding, tolerance for failure), funds AI and biotech safety as seriously as capability, and treats its most cost-uncertain element — high-speed rail — with discipline rather than romance. Done that way, it is not spending so much as compounding: an investment in the basic knowledge and human capital from which the next century of American prosperity, security, and flourishing will be built.

Sources

← Back to The Great Reinvention