How to Build Upon Vannevar Bush’s “Wild Garden” to Cultivate Solutions to Human Needs

Wednesday, July 14, 2021

Issues in Science and Technology

Like many brilliant researchers, Vannevar Bush had a knack for raising the right questions. As we commemorate the anniversary of his groundbreaking Science, the Endless Frontier report and consider its lessons for us, we should be guided by Bush’s forward-looking, probing questions at least as much as by his more time-bound answers. Many of Bush’s specific proposals were altered or rejected outright, and some that were implemented are ripe for reassessment, but his questions remain as pressing as ever.

Bush started with a broad, fundamental query: What can be done to make the US population healthier, safer, more prosperous, and more comfortable in the postwar era? Then he zeroed in on the relevance of the areas he knew best and asked: What US research institutions and funding modes should be created, restructured, or adapted to enable the nation to succeed in the very different world created by the war? 

Bush did not ask any questions that were premised on protecting the status quo. He advocated fresh thinking, not orthodoxy; when the report made waves in academic and political circles, he did not shy away from the controversy. It is only in retrospect that we fully appreciate his vision’s enduring value.  

In considering the challenges of our moment, Bush’s farsighted questions about both outcomes and means should inspire us to take a clear-eyed view of the US research enterprise and to propose new approaches to confront the most pressing issues in today’s world. Happily, unlike Bush, we do not have to build a system almost from scratch. But we also cannot simply take refuge in the comforts of the status quo and repackage familiar arguments for doing more of the same.

Compared with 1945, our current geopolitical circumstances are vastly different. Bush wrote, in part, because the United States could no longer rely on science coming from a shattered Europe and he saw that the nation had an unprecedented chance to become the world’s dominant scientific and industrial power after the war. Today, we face growing scientific and economic competition, particularly from a rising China that is also a military and ideological rival. Regardless of how one views China, its government, or its ambitions, there should be agreement that this new dynamic represents a fundamental challenge to the United States. 

To meet this moment, we need to ensure that our federally sponsored research addresses questions that will enhance our competitiveness now and in the future. At the same time, we need better ways to usher more of those research advances into the marketplace. Our current system has many strengths—top research universities, a thriving basic research enterprise, an entrepreneurial ethos, prospering venture capital—but we must not allow these historical advantages to blind us to gaps that could become fatal weaknesses.


The United States has at least two areas of weakness that require attention. The first is that we have largely abandoned civilian use-inspired basic research: fundamental research specifically designed to solve practical problems. Organizations such as Bell Labs were once known for this type of long-term but targeted research. Most of the big industrial labs that performed this research have effectively vanished, and universities (and their federal funders) have tended not to pick up the slack. Instead, universities and federal agencies focus, by and large, on curiosity-driven basic research: research that is essential to our nation’s long-term success, but that by itself is not sufficient to make us competitive in today’s world. 

What is use-inspired basic research? The Bell Labs work that led to the creation of the transistor and the semiconductor is perhaps the premier example. On the one hand, the research probed fundamental questions in physics and was recognized with a Nobel Prize. On the other hand, the research was targeted toward a specific, practical goal: replacing vacuum tubes, which were unreliable and too energy intensive. Our economy continues to benefit from Bell Labs’ decades-old breakthroughs—but we cannot live off that technological inheritance forever. 

There is no shortage of use-inspired questions we should be pursuing today. One example is developing algorithms that would allow computers to “learn” using less data. Our current data-intensive algorithms put the United States at a competitive disadvantage; China has access to more data than we do because of its larger population and its weaker privacy protections. One way to develop algorithms that need less data is to study how young children learn—an important and difficult basic research problem—with an eye toward giving computers some of that capability. Unlike computers, children do not need to see, say, a thousand cats to properly identify a cat. If we could make our computers’ need for data similarly economical, we could vault past the competition.  

In a competitive environment, we cannot assume, as Bush did, that curiosity-driven advances in science someday will be useful in some fashion. Instead, we must also be pursuing questions targeted at needed breakthroughs. Bush envisioned science as a kind of wild garden: individuals seeding ideas based on their intellectual interests with no overall design. We also need to see science as a kind of farm, where people work together to cultivate and advance selected ideas to address human needs. These two types of science need each other in order to advance and for our nation to thrive. Our current system does not provide the optimal balance between them.  

The second gap is that the United States is not as good as it needs to be at getting “tough tech” ideas to market. Tough tech encompasses new products and processes that involve hardware as well as software—the development of which could create whole new fields. Such products can also help address pressing problems, such as climate change. Tough tech is often too risky and takes too long to mature to be appealing to angel investors, venture capitalists, and others who finance new companies. At the same time, the work is too far along the research and development path to qualify for government support. The result is that we miss the boat. Sometimes these companies instead get either financed by foreign investors or finally get developed and manufactured overseas (or both)—or they simply fold.

As a university president, I hear all the time about promising ideas that struggle for air in our current system. For example, there is the case of a professor with a brilliant new approach for grid-scale energy storage, but the path to commercialization would stretch beyond the typical five-year limit for venture capital funding. Or there is the potential approach to treating Alzheimer’s disease that could not attract funding to get beyond the lab because of the extended timeline to develop the treatment.

It is no simple task to figure out how to create a system that would provide the funding that tough tech ventures need to survive. But we make it more difficult when ideology or hardened views about institutional roles are used to rule out proposals before they can be examined. Throwing around the label “industrial policy” is not a serious mode of analysis, regardless of whether the term is being used as a blessing or a curse. 

Government, industry, and academia will have to work together to create new mechanisms to sustain tough tech companies long enough that their ideas can be fully developed and scaled. One approach we’ve launched at the Massachusetts Institute of Technology is setting up an independent entity we call The Engine, which provides technical assistance, space, and capital to support tough tech founders. But that’s just one effort in one location. The federal government should be designing tax and funding policies to help more places experiment with more ways to get new companies off the ground.

These gaps in our system may seem like mere frustrations or peripheral problems now, but they can grow into fundamental failures, especially when other countries are working to benefit from our shortcomings. 


As a nation, we have the wealth, the talent, and the creativity to get ahead of these problems; what we have lacked is the will. That is why I have been such a vocal supporter of the bipartisan, bicameral bill that is aptly named the Endless Frontier Act, which takes aim at some of the issues that I have discussed in Issues previously. The Senate passed the bill in June 2021 as part of the US Innovation and Competition Act to increase “investments in the discovery, creation, and manufacturing of technology critical to US national security and economic competitiveness.”

Although the Endless Frontier Act is far from perfect and should be further refined, it gets the fundamentals right, and those have emerged from the Senate process intact. The bill would create a new Technology and Innovation Directorate at the National Science Foundation (NSF) that would be responsible for funding the use-inspired basic research our nation needs across a wide range of fields, and for related educational, tech transfer, and test bed activities. The measure would also authorize significant new funding for the effort: the new directorate’s annual budget would reach $9.3 billion in fiscal year 2026, the last year of the bill. 

NSF is the appropriate agency to anchor this effort: its portfolio has great breadth, the agency has deep experience working with universities as well as relationships with industry, and it has a deserved reputation for excellence. Beyond that, NSF works closely with universities, and universities educate students as an integral part of carrying out research. In the end, the source of a nation’s strength is its people, and educating students in the latest technological areas is an essential part of becoming more competitive as a nation.

The House is also moving forward with an important and valuable effort through its National Science Foundation for the Future Act. This bill is more notable for its similarities with the Senate approach than for its differences. The NSF for the Future Act also creates a new directorate with new funding to address competitiveness issues (among others) through use-inspired and translational research. Both bills allow and encourage—but do not mandate—the new directorate to experiment with different project selection processes, and to experiment with different hiring practices to bring in well-regarded experts to run the new programs. The Senate bill, by comparison, explicitly mentions the Defense Advanced Research Projects Agency (DARPA) as a possible model for some programs; the House bill does not call out DARPA but gives NSF similar flexibility in implementing its research programs. The House bill permits all the directorate activities laid out in the Senate bill.

So then what are the issues that will need to be ironed out between the House and the Senate? Although this is not an exhaustive list, here are some matters that will require attention. First is determining the scope of the directorate, including what broad issues should be within the directorate’s purview, what kinds of research it should fund, and what technologies it should focus on. The House bill lists six societal issues on which the directorate can focus, including competitiveness; the Senate bill is focused primarily on competitiveness. 

Competitiveness, I believe, is an overarching concern that will affect the nation’s ability to deal with climate change, equity, and other important societal issues listed in the House bill. It is also an arena that is a logical extension of current NSF concerns and activities. Working on issues beyond competitiveness may be a worthy idea, if funding is adequate and the directorate can still be sufficiently focused. At the very least, competitiveness should be an initial priority for the new directorate. 

Both bills could be clearer on what flavor of research the directorate will fund, while still giving NSF latitude to make scientific decisions and to adapt its plans to new or evolving needs. As already noted, at least one focus for the directorate should be funding long-term, use-inspired basic research targeted at problems with practical implications. The solutions to such problems typically demand teams with crossdisciplinary expertise. 

Finally, the Senate bill includes an initial list of key technology focus areas, including broad topics such as artificial intelligence and quantum information sciences. Congress frequently provides this type of guidance to NSF and other agencies, and this guidance can be helpful in getting programs off the ground, as well as preventing misunderstandings and squabbling. The specific list in the bill might be refined or shortened, but it certainly should not be expanded; focus is of vital importance. The key technology focus areas will need to evolve over time, but they should not change frequently, as it will take sustained research in a focus area to make meaningful progress.

The House and Senate will also have to work out how to address racial and geographic equity. Both bills appropriately have provisions intended to attract more individuals from underrepresented minority groups into science and engineering, as well as to assist schools such as historically Black colleges and universities that have large enrollments of students from these groups. These provisions are not only a longstanding matter of equity and justice; the United States will never be as strong and prosperous as it should be without drawing on the talents of all segments of our society. 

Both bills also create new programs to help build the capacity of so-called emerging institutions—schools that participate in federal research programs, but that could expand the kinds of research experiences the schools provide to faculty and students. Some of these emerging institutions also have significant minority enrollments.

Such efforts, while focused on racial equity, will also serve to widen the geographic distribution of research funds. But more than this effort is needed. The geographic distribution of research funding was a subject of debate at NSF’s founding, and the issue remains thorny.

The United States cannot reach its full potential if research and the resulting economic activity are confined largely to the coasts. Such uneven distribution also limits opportunities for many Americans. The question then is how to increase capability around the nation without diluting the focus on excellence and the concentration of resources that have enabled the creation of Silicon Valley in California, Kendall Square in Massachusetts, and other success stories. 

The Senate bill includes important provisions to create economic winners around the country, perhaps most notably a Commerce Department program that would create regional technology centers and offer regional planning grants. The goal is to help move ideas from universities around the country into commercial production. Research work from anywhere in the country might have applicability that could spur growth in regional industries. 

Other Senate provisions are more controversial. The goal of providing money that helps all regions is an appropriate one. But Congress needs to come up with a distribution approach for research funding that raises all boats, without inundating some boats with more water than they can handle or leaving some premier vessels high and dry. Congress should negotiate a way to build research and training capacity at more universities, without spreading money so widely or with so little focus that overall US capacity is weakened. No region of the country will benefit if the United States falls behind our competitors in research and technology due to funds being spread too thinly to make a difference. Likewise, the nation will be damaged if only a few regions prosper.

A final unresolved area of concern is national security. Congress rightly wants to be sure that the United States will be the prime beneficiary of taxpayer investment in research, and it’s particularly concerned about China taking undue advantage of US researchers. Universities have been increasing their own scrutiny of engagements with Chinese entities. In 2019, for example, MIT set up a new review process for research engagements with China, Russia, and Saudi Arabia, looking at their potential effect on US national and economic security as well as on human rights.

But we cannot become so concerned about standing in the way of China that we hurt our own ability to move forward. In the end, the most successful strategy for competing with China is having a robust research and technology system in the United States. Policies that create costly barriers to research—by unduly hindering open research, by making it harder to attract the best graduate students from around the world (most of whom remain here after completing their PhDs), or by alienating those who want to support US universities—will do little to hinder other nations but a lot to hurt ours. 

These assertions do not mean that no additional policies are needed. Reasonable requirements to disclose foreign gifts and contracts; prohibitions on faculty engagements that create conflicts of interest; targeted limits on sharing research in narrowly defined, problematic areas—all these areas can protect the United States with sound, clear policies that provide useful guidance to universities. But if we impose broad measures that simply send the message that we are cutting ourselves off from the talent and ideas the rest of the world has to offer, we will limit our own ability to move forward. The worst situation for our national security would be producing nothing that anyone thinks is worth stealing.

You don’t win a race by expending all your energy on tripping up your opponent. Rivalry can spur innovation, but fear is paralyzing. Vannevar Bush was uneasy about the future, but his report’s power came from his quiet confidence in US potential. We need to draw on that. We have plenty of cause for confidence. If we falter, it will be because we have become too set in our ways, too complacent, too unwilling to ask provocative questions, too worried about others’ successes and failings and not enough about our own. Instead, we need to build on our strengths.

Bush did not choose the name for his report lightly. He understood that seeking out frontiers takes gumption, vision, energy, confidence, and daring. In the end, his most important message to us may be to continue to embody those attributes.