Introduction

IntroductionWe live in an age of horrors and wonders.Wildfires, hurricanes, tsunamis, floods, heat domes, droughts, rising oceans, melting ice caps and glaciers, thinning water tables, shrinking biodiversity, and ubiquitous pollutants all chide us for our climate delinquency. A pandemic tests our medical and political systems. War, racism, poverty, coups, dictatorships, fanaticism, partisan polarization, crime, and refugee surges mock and batter our democratic and humanistic ideals. Our mental health declines. Expected lifetimes and learning shrink. And global tensions bring heightened risk of nuclear catastrophe center stage once more.Despair is one response. The other? Let’s get to work.The latter prevails among the talented and curiosity-driven intellects that power the vast domain of research and development. Universities play a small but disproportionately important role in that enterprise because they perform the lion’s share of basic research that advances understanding of our world. Science, technology, engineering, and math scholars pave the way for the next scientific breakthroughs, while their counterparts in the social sciences and humanities illuminate the promise and peril of human ingenuity in a period of accelerating technological tumult.The ultimate results, and pace of progress, can be breathtaking.Covid-19 vaccines, based on decades of mRNA research, have saved, according to one study, 3.2-million lives and averted 18.5-million hospitalizations. A 20-foot-long spacecraft took a half-pound bite out of an asteroid, dropped the sample of primordial materials off on Earth for analysis, then continued on its journey after already traveling four-billion miles. The computational power of cellphones is 5,000 times faster than 1980s supercomputers and 900-million times faster than the Apollo 11 guidance computer. Precise sensing instruments gather new evidence of gravitational waves, with implications for the very nature of time and space and the origins of the universe. Prototype pilotless AI aircraft fly in collaboration with fighter pilots. AI programs spot early cancer that doctors don’t.At the same time, after decades of scientific cooperation, tensions have ratcheted up perilously between the United States and China, the world’s two most productive research powers. That antagonism not only undercuts collaboration among great thinkers, it also hinders cooperative solutions to global problems. Further, such geopolitical tensions push research into applied realms and away from the kind of basic research that many scientists say might, in the long run, have more profound and lasting benefits.The goal of this report is for readers to better understand the scope of the American academic-research enterprise, with emphasis on STEM fields because those are where funding streams and federal legislation exert the most influence. This report discusses how that enterprise fits into the international R&D puzzle. It tries to capture the immense intellectual passion and imagination of research scholars and the values that guide them.Through examples and insights from scores of experts, readers will learn:how universities are trying to balance basic, applied, and developmental research, where support for that work comes from, and who is performing it;why tensions between the United States and China are threatening more than four decades of research collaboration and helped spur the Chips and Science Act of 2022;how the incentives, prestige, and mystique around top Carnegie classifications and membership in the rarefied Association of American Universities influence higher education;what types of partnerships universities are forming with government and industry;and how exciting prospects in four particularly intense research fields — sustainable energy; robotics, biotech, and health; artificial intelligence; and quantum science — are pushing the frontiers of discovery.With that information, readers will be able to strategically align their institution's research endeavors with its strengths and mission.Return to Table of ContentsSection 1: The Vast R&D EnterpriseAmerican colleges and universities spend almost $90 billion a year on research and development. We take for granted that universities will apply vast resources and the intellectual talents of their faculty to the most daunting challenges of today and the most vital quests for tomorrow. But it was not always so. Only within the last century did war, international competition, and societal need gradually generate American academe’s colossal research edifice. And it began largely through the efforts of one man.A “quintessential Yankee, a politically conservative Republican, at times prickly, elitist, even arrogant.” So Sol Gittleman, a longtime former provost of Tufts University, describes Vannevar Bush in his new book, An Accidental Triumph: The Improbable History of American Higher Education.Although Bush’s name “is unrecognizable to the vast majority of educated Americans,” Gittleman explains, “he is responsible in large degree for the enormous research achievements of American universities, as well as the success of our higher-education establishment in general in the 20th century and the continued worldwide leadership of American universities going into the third decade of the 21st century.”Bush earned a bachelor’s and a master’s degree in electrical engineering from Tufts, and a doctorate in engineering jointly awarded by Harvard University and MIT. It wasn’t just his technical grounding, however, but Bush’s “organizational agility” that enabled the “Godfather of post-war Big Science,” through government, foundation, and university channels, to become a prime force behind the Manhattan Project and, later, a burgeoning research network among government, academe, and industry.In 1945 Bush sent President Harry S. Truman a manifesto, Science: The Endless Frontier, which set out an ambitious blueprint for U.S. research and fostering research talent, to bolster health, defense, and the public welfare.Until World War II, government support of academic research was minimal. That began to change in 1941 with the Office of Scientific Research and Development and its successor, the National Science Foundation, established in 1950.The five-year gap “between Bush’s proposal and the start of the new agency gave other federal agencies — notably the Atomic Energy Commission, the National Institutes of Health, and the Office of Naval Research — the opportunity to expand their portfolios,” France A. Córdova, then director of the NSF, wrote in 2020, the 75th anniversary of Bush’s manifesto and the 70th of the NSF. “Thus, the new NSF did not become the funder of all basic research across the federal government that Bush had envisioned. Yet in many ways, the report had the impact Bush sought, and his influence on science policy continues to be enormous.” In its first year, NSF’s appropriation from Congress was a paltry $3.5 million. It wasn’t until 1957 that the Soviet launch of Sputnik lit a fire under policymakers, who the next year more than tripled NSF funding, from $40 million to $134 million. In 1969 Congress allocated almost $500 million for NSF projects. And NSF’s budget has long been eclipsed by that of the National Institutes of Health, which, by 2021, had an annual budget of $45 billion.Big Science, Big MoneyWe’ve glimpsed how the funding seeds that Vannevar Bush and like-minded visionaries planted in the middle of the 20th century bloomed into an astonishing research realm. And the NSF and NIH are, of course, only two of many federal agencies and industry sources supporting today’s university-based research. Moreover, university-based research is itself a relatively small, though vital, piece of a much larger national R&D picture.Now let’s take a closer look at today’s research landscape. In doing so, we are going to focus mostly on STEM fields because federal support for the humanities is dwarfed by R&D spending on STEM. The 2024 budget for the National Endowment for the Humanities is $211 million. Colleges, universities, and philanthropic groups will make up the rest. But in recent years, the total from all sources has amounted to less than 1 percent of what’s spent on STEM research and development. The National Center for Science and Engineering Statistics, which is part of the NSF, publishes yearly data from the Higher Education Research and Development Survey. The HERD Survey’s 2022 statistics on science and engineering research convey the scope of a robust academic enterprise.Buckle up for some numbers as we highlight a few trends in scientific-research funding:University R&D spending is up. Academic institutions spent $89.9 billion on research and development in 2021, $3.4 billion more than the year before. Federal support accounted for $3 billion of that increase, a jump not seen since 2011, when the American Recovery and Reinvestment Act, passed during the Great Recession, supplemented traditional R&D channels. State and local government support of university research in 2021 increased by $149 million (3.2 percent) over the year before, but that was canceled out by a $150-million (2.6 percent) drop in support by nonprofit organizations.The U.S. government supplies most university-research support. The federal government provided 55 percent of academic-research support in 2021 — $49 billion. Universities themselves fund about a quarter of their R&D. Nonprofits, industry, and state and local governments make up most of the rest.In federal support, the Department of Health and Human Services, which includes NIH, funded $27.5 billion, mostly in biological, biomedical, health, and other life sciences. The Department of Defense funded $7.4 billion. NSF: $5.4 billion. The Department of Energy: $2.2 billion. The National Aeronautics and Space Administration: $1.8 billion. The Department of Agriculture: $1.3 billion. Other federal sources combined: $3.5 billion.Research is integral to how universities function. It is part of the three-legged-stool, along with teaching and service, for faculty careers, a model built into the American higher-education tradition and most tenure-and-promotion systems. Research experience, from undergraduate years up, has been increasingly valued by students, colleges, and employers alike. There are 146 “high intensity research activity,” or R1, universities in the Carnegie Classification system, and the prestigious Association of American Universities has 71 member institutions. (More on those in Section 3 of this report.) The NSF has also established programs and incentives for higher-education institutions collaborate with one another and with regional industries and governments to further spread and democratize R&D resources and know-how.However, in terms of research expenditures, the greatest might is concentrated among a few dozen power players. In 2021, 30 institutions accounted for 42 percent of total R&D spending within the higher-education sector. Twenty-four spent at least $1 billion. Sixteen of those top 30 were public institutions that spent a combined $20.2 billion, while 14 were private universities that spent a combined $17.9 billion. Unsurprisingly, 27 of the top 30 have medical schools. At the top of the spending list, at $3.18 billion, is the Johns Hopkins University, which has not only a large medical network but also the huge Applied Physics Laboratory, which we’ll look at more closely later in this report.U.S. R&D spending is big, but international R&D spending is enormous. The National Science Board establishes policy for the National Science Foundation and also advises the president and Congress. The board issues an invaluable annual series of reports on science-and-engineering activity. Data from those show that U.S. research-and-development performance (the research and development itself, from all sources, plus its planning and administration) reached $708 billion in 2020. Global R&D expenditures reached an estimated $2.4 trillion in 2019.In the United States, businesses are the major players, accounting for 75 percent of R&D performance in 2019, while higher-education institutions and the federal government accounted for 12 percent and 9 percent, respectively.Growth in American R&D performance has been impressive, rising an average of 3.8 percent annually from 2010 to 2019 while gross domestic product over the same period averaged 2.2 percent a year.Wow, right? Except that …China’s R&D performance is growing at almost double the rate of America’s. So while in global R&D performance the United States led China 28 percent to 22 percent in 2019, it might not maintain a lead for long. Those numbers reflect a broader gradual shift in R&D momentum from the United States and Europe to East-Southeast and South Asia. That, alongside other geopolitical tensions, helps explain the increasing scrutiny in recent years of American cooperation and collaboration with science and technology.Does Basic Research Get Its Due?The Science Board figures also point to another source of chronic tension within American R&D: how to balance basic research with applied and experimental-development research. That is, research for the sake of advancing fundamental knowledge versus research increasingly directed toward specific goals and then products and processes.As we’ve seen, in the United States, colleges and universities spend about $90 billion a year on R&D. That is only a little more than a 10th of overall American R&D spending, and less than 4 percent of the global $2.4 trillion in R&D spending.How is it, then, that so many important discoveries like the structure of DNA and the role of mRNA have emerged from academic labs in the United States and abroad, often through transnational academic-research collaborations? Why, filtered through the 5.14-million research articles published yearly in almost 47,000 academic journals, do so many key findings emerge from universities, or at least scientific teams that include universities, when they account for only a small part of the research domain?Because while academic researchers are responsible for only a small fraction of research, they account for the largest fraction of basic research.Consider:In the United States, NSF data show, academic institutions performed 48 percent of basic research in 2019. The business sector performed about a third, with federal laboratories and nonprofit organizations performing 12 and 9 percent, respectively.Almost two-thirds of the R&D performed by academic institutions is basic research, roughly a quarter is applied research, and about one-tenth is developmental research.The federal government funded about 41 percent of all basic research performed. The business sector performed 32 percent and funded 33 percent. In 2019, basic research comprised $102.9 billion (15 percent) of total R&D spending, compared with applied-research spending of $132 billion (20 percent) and experimental-development spending of $432 billion (65 percent).OK, then, while basic research doesn’t get the same support as applied and developmental research, all types of research get what sounds like a whole lot of money. What, then, is causing tension?Incremental trends. They worry some experts the same way seemingly small but gradual economic trends worry analysts from Wall Street, the Federal Reserve, and the Treasury Department, who spy cracks in the system that augur bigger problems. Incremental changes over years and decades become, or reflect, significant shifts in priorities.Between 2010 and 2019, while America’s annual basic-research spending rose about $12 billion, its share of such funding declined — from 19 percent in 2010 to 15 percent in 2019.Upstream and Downstream DiscoveriesThe tension between research for its own sake and research in service to what seems an important goal is an old story. In fact, as Sol Gittleman recounts in An Accidental Triumph, it was Vannevar Bush’s uncompromising defense of researchers’ independence that alienated congressional leaders in the late 1940s and delayed and stunted the birth of the NSF. As they often still do, legislators wanted more say and oversight as to where that research money would go.And while particulars have changed, there is a Cold War-style framing to the discussion now, with China and Russia looming large as U.S. competitors and/or antagonists, just as there was after World War II (although Russia was then seen as the more immediate threat). The Chips and Science Act of 2022, the America Competes Act of 2007 and its reauthorization in 2010, and similar federal pushes dating back to the Sputnik-spurred National Aeronautics and Space Act of 1958 have often been compelled by fears over America’s technological, economic, and educational standing and self-sufficiency.When hundreds of billions of dollars are concerned, a few percentage points matter, and the current direction is clear.“We are pivoting slowly away from science-based discovery,” says Ishwar K. Puri, senior vice president for research and innovation and a professor of aerospace and mechanical engineering at the University of Southern California. At some point, he says, the change will be felt in a significant way. “There will be a tipping point.”The funding pressures push researchers toward more project-oriented, interdisciplinary team efforts. That entails tradeoffs, says Puri. Researchers are creating real solutions to real problems like climate change and dementia. But, he asks, “will we neglect to pay attention and properly fund the upstream, science-based discovery?”Discovery is built on discovery: the fundamental studies of virology that Jonas Salk built upon to create a polio vaccine, the pioneering electrical engineering that preceded the invention and improvement of the integrated circuit, the core understanding of the immune system leading to Jim Allison’s breakthrough immunotherapy treatment for cancer.Was the space race the right obsession for the 1960s? Are semiconductors a sensible preoccupation today? Quite possibly.But, says Puri, “It’s this issue today and that issue tomorrow.” If goal-oriented dictates pull the best research minds from their utmost interests and passions upstream, what might be the costs decades from now in downstream advances that did not come to fruition?“Of course that worries me,” says Holden Thorp, editor-in-chief of the Science family of journals, about the drift away from basic research.“In biomedical research, lots of things work in vitro, some things work in cells, fewer work in mice, and only a precious few actually work in humans. There is also a huge gap between basic research and immediate application, known as the ‘valley of death,’ where innovations get stuck,” Thorp wrote in a Science editorial.The federal research complex is good at science but less good at turning discoveries into medicines, says Thorp, a chemist, former chancellor of the University of North Carolina at Chapel Hill, and former provost of Washington University, in an interview with The Chronicle. Pharmaceutical companies that take the discovery handoff and run with it sometimes have more success, as with the Pfizer and Moderna Covid vaccines.Also, says Thorp, biomedical research is slow. The dementia drugs coming to market now are based on several decades of bench science and corporate drug trials. Aiming federal funds at a disease sets the public up for disappointment if people think progress comes in Capitol Hill funding cycles of several years.And, Thorp says, setting research agendas in Washington can also skew priorities. Many in the academic-research world would like to see NIH funding remain robust but basic-science funding at NSF increased, he says. Wars on disease are political winners — lawmakers, like the rest of us, all know folks who have died from cancer or suffered from Parkinson’s. A new particle-physics project is a tougher sell.The shift toward more applied and developmental research isn’t random budgetary drift, it’s very intentional, says Shalin Jyotishi, senior adviser for education, labor, and the future of work at New America, a Washington think tank. “There has definitely been a strategic shift in what kind of R&D the federal government will fund,” says Jyotishi, who has also served as a staff member or adviser for the NSF, the World Economic Forum, and a number of other science policy and governance organizations.No one in respected policy circles is foaming at the mouth and demanding Soviet-style five-year industrial plans. But the mood started changing in the 1990s, he says, and shifted more significantly several years ago, right around the 75th anniversary of Vannevar Bush’s Endless Frontier.“If I were to point to a singular theme,” Jyotishi says, “I would likely call out competition with China. … I really think that has been a clarion call among advocates of use-applied research.”And, he says, “I think this is the right direction.”Jyotishi is skeptical of canonizing some purist notion of upstream basic science outside societal priorities. Serious funding of the NSF itself was a response to Sputnik. The internet evolved partly from the military and academic Arpanet, allowing computers to communicate with one another away from mainframes. GPS arose from satellite-based monitoring of missile-carrying submarines. The discovery of Vitamin D was a response to the widespread misery posed by rickets. And so on, says Jyotishi. “There needs to be more emphasis on translation” — the term of art for conveying knowledge into usable application.It’s not unlike what other countries do, he says. Israel, Singapore, Japan — most OECD countries with robust R&D funding think about how to make the most of government research and money.As does China.Return to Table of ContentsSection 2: China and ChipsIn recent years, the United States has increasingly viewed China as a competitor and a threat. That perception shapes America’s foreign policy as well as its economic, industrial, research, and educational priorities.Although not generally stated so explicitly, the Chips and Science Act of 2022, a decade-long investment of $250 billion — presuming that money is actually appropriated — is intended in part to decrease the United States’ dependence on Asia and especially America’s vulnerability to a chokehold on tech imports from Taiwan if it is invaded or blockaded by China. Such strategic scenarios are complicated, but the act’s goal is to boost America’s advanced-microchip production and its research-and-development performance.In addition, the Biden administration has imposed stringent restrictions on sharing certain microchips and chip-manufacturing equipment with China. Under serious discussion are further restrictions on American investment in development of sensitive, advanced technologies within Chinese borders. American business confidence in China is already slumping in any case, and detentions of American executives, through “exit bans,” aren’t helping.Some critics doubt whether the Chips and Science Act will have the desired results and whether America’s chip-export controls are effective. They even suggest that such controls might be self-defeating because they could spur China to ramp up its own chip production and technical advances faster than it might have otherwise. These critics argue that stringent barriers to more specific technologies would make better sense.Meanwhile, as policymakers try to sort legitimate research-and-development fears from possible paranoia, the science and technology agreement, or STA, that the United States has had with China since 1979 hangs in policy purgatory. The State Department announced in August 2023 that the agreement is receiving a six-month extension, which an article in Foreign Policy refers to as “life support.”The 11-article agreement, signed by President Jimmy Carter and Deng Xiaoping when diplomatic relations were normalized, provides for reciprocity in research and development, ease of scientists’ travel between the countries, cooperation in publishing research results, and a joint oversight commission. A 2018 amendment addressed issues of intellectual property.The recent tensions have led to serious concerns about the agreement: Does China’s ascendancy since 1979 make the pact less meaningful, or more? Either way, is China a trustworthy-enough partner to maintain such an arrangement?Research and development that can be applied to self-guiding weapons and weapon-delivery systems, cyberattacks, and biological weapons are among many areas of near-term concern. Looming large in the longer term are potential advances in quantum computing, particularly its encryption and decryption capacity.Ratcheting up tensions are China’s tacit backing of Russia’s Ukraine invasion; a lack of cooperation and transparency throughout the Covid-19 crisis; human-rights abuses against the Uyghur population and other minority groups; China’s unnerving collecting of worldwide DNA samples; intellectual-property theft; AI-driven disinformation campaigns; hacking; its detention of several Americans; and alleged spying. Nor are minds eased by China’s military and nuclear-warhead buildup; its aggressive patrolling of international and contested waters, sometimes with Coast Guard vessels armed like battleships; its dangerously harassing U.S. surveillance flights in international air space; the mysterious disappearance of its defense minister and other top officials; and its 2021 testing of a hypersonic missile that travels at 20 times the speed of sound and can reach anywhere on earth in less than an hour.The press office of the Chinese embassy in Washington didn’t respond to Chronicle emails or phone calls seeking comment on the STA. But China has urged renewal of the agreement. “As two major R&D countries, China and the United States should maintain contact and exchanges” in science and technology, a Chinese embassy spokesperson told Science in 2023.More generally, in the wake of the Hamas terrorist attack on Israel in October 2023, which destabilized global politics even further, Xi Jinping, China’s president, sought to play down differences with the United States. “We have 1,000 reasons to make China-U.S. relations work well and not a single reason to make China-U.S. relations bad,” he told Sen. Chuck Schumer of New York, who was visiting China with a congressional delegation. A little more than a week later, however, Xi was rolling out the red carpet for Vladimir Putin.China also has retaliated for U.S. chip-export restrictions with its own bans on iPhone use in government offices and on the export of minerals and metals used in the production of semiconductors, missile systems, and solar cells. And if distrust of China is rampant in the United States, the feeling is mutual, with China launching a massive antispying campaign that, The New York Times reports, “blurs the line between vigilance and paranoia,” including mandated courses for university faculty members “on protecting state secrets, even in departments like veterinary medicine.”The STA was established in part to help modernize and stabilize China, to assist it in reaching its economic and scientific potential. Its R&D performance is now second only to that of the United States, and in recent years, as detailed in Section One of this report, China has been out-pacing America’s R&D growth. China’s economy, however, is experiencing severe turbulence, with stalled economic growth, a collapsing real-estate bubble and domino effects in the investment-banking sector, and a young-worker jobless rate that reached 21.3 percent before China stopped reporting it.End of an Era?Given this turbulent backdrop, should the STA be renewed? Could almost a half-century of research cooperation really end?One scholar who has worked extensively in China hopes not, but worries that “it may be hard to move forward as things stand.” Until recently, Denis Simon was a clinical professor of global business and technology at the University of North Carolina at Chapel Hill. Before that he held high-level positions in academe and consulting, including the executive vice chancellorship of Duke Kunshan University, in China.He says he resigned from his UNC position during the summer of 2023 after the provost, Christopher Clemens, chided Simon for hosting a meeting that included UNC colleagues and a delegation from the Chinese embassy. Simon has arranged such meetings, and traveled to China, for more than 40 years, he says in a phone interview. (In a later column for China Daily, he details the UNC clash.)The Chinese delegates weren’t seeking sensitive technical information or operational details about the university, says Simon. There was no press or publicity. And, “to be blunt,” he says, “it was extremely productive.” The lack of communication with China, he adds, “is one of the major reasons we’re at loggerheads with one another.”He says that Clemens nonetheless instructed him to desist from hosting such meetings unless Simon sought and received formal permission from the UNC academic leadership. “The more I thought about the hard slap on the wrist that I got, the more I thought this is not the place for me,” says Simon. Earlier in 2023, before traveling to China, Simon says he was heavily vetted by UNC security, export control, global affairs, and research officers.In response to a Chronicle query to Clemens’s office about the situation, the university issued a statement saying that it could not comment on Simon’s resignation because of state human-resources law on personnel matters. But the statement goes on to say:“UNC-Chapel Hill has a steadfast commitment to maintaining the integrity of research conducted by our faculty, staff and students. We seek to identify and manage any potentially conflicting relationships to preserve transparency and independent decision-making. At the same time, we take very seriously legitimate concerns about the need to safeguard U.S. academic research from improper foreign influence in the interest of national security.”Simon, now a consultant who has secured a new book contract, says the clampdown on him is emblematic of the widespread paranoia about China that is gripping American campuses. That the STA hasn’t been renewed further illustrates that political polarization around China has hopelessly infected the research realm, he says.That despite the fact, he says, that the Justice Department’s so-called China Initiative to root out spying in academe was terminated after yielding few charges and establishing a chilling civil-liberties atmosphere across campuses, especially for Asian Americans.Similarly, the Trump administration and some lawmakers considered Confucius Institutes, the more than 100 language and culture centers based in the United States and cosponsored by the Chinese government, a threat to national security. A review under the auspices of the Department of Defense found no evidence of that. The institutes’ stance toward academic freedom was a different and possibly more credible concern. Most of the institutes closed during the review process, although a report from the National Association of Scholars says the centers have been springing back to life under new names.“They were an instrument of Chinese soft power, no doubt," says Simon, "but not dens of espionage.”Simon acknowledges that American universities, in their dealings with China, gradually “took our eye off the ball.” New safeguards for intellectual property should be negotiated. But, he says, the vast majority of Chinese scientists who come to America to study and work simply want to advance their careers and their fields. Those who return to China stay plugged in, in a constructive way, to American progress and values. Those who remain in the United States provide much-needed talent that it can’t afford to lose.The irony, says Simon, is that the relationship is deteriorating just as China has advanced far enough to provide invaluable collaboration in dealing with future pandemics and other public-health threats, as well as climate change, clean energy, food security, Alzheimers and other brain research, as well as other key global challenges. That collaboration “could be the magic elixir,” he says, and without it America is forgoing the rewards of a relationship in which it has invested for decades. The current political climate is also dissuading a generation of young American students from learning about China, another waste of talent and interest that will make it that much harder to mend relations in the future, he says.The Trust FactorIs the relationship salvageable at this point?The Justice Department’s China Initiative was shut down because of the civil-liberties chill, not because the FBI decided everything was OK. The department decided that the combined threat from China, Russia, Iran, and North Korea requires a broader approach. And officials remain astonished at the scale of Beijing’s efforts to steal secrets, technology, and research. In early 2022, according to a National Public Radio report, the FBI said it was handling 2,000 China-related cases, with a new one being opened every 12 hours.“Given what we now know about Chinese ‘talent recruitment’ and espionage programs” involving both Americans and Chinese nationals, the risks of collaboration outweigh the benefits, says Dan Blumenthal, a senior fellow at the American Enterprise Institute, a Washington think tank, in written responses to The Chronicle’s queries.One would have hoped that biomedical concerns, at least, could be fenced off from geopolitical competition, says Blumenthal, who previously served as senior director for China, Taiwan, and Mongolia at the U.S. Department of Defense. But China's “terribly irresponsible behavior” around Covid-19 and “its extreme ‘medical nationalism’ should give us pause about what might happen if China makes a breakthrough in this field before the West as well,” says Blumenthal. “Will it share the benefits? It is a real question.”The Confucius Institutes “ended up not being much of a threat,” he says, and “there is nothing intrinsically wrong with China spreading its culture. Chinese civilization has made important contributions to the world.“What we now know, however, is that the Chinese Communist Party, not the Chinese people, is interested in malign influence over governing, cultural, and business elites. And that is what we have to guard against.”China’s economic problems, too, are reason for greater vigilance, not more cooperation, in Blumenthal’s opinion. President Xi Jinping, Blumenthal says, “has sent China into a self-inflicted economic crisis by cracking down on the most productive sectors of his economy. He has also made the environment for cooperation much more difficult. The general political environment is far more totalitarian than it was even just a few years ago.”Universities that have a campus in China should examine their options and those contemplating opening one should think long and hard, says Blumenthal.“The Chinese government runs the largest state-sanctioned intellectual property-stealing programs the world has ever known. All university centers setting up shop in China need to be aware of this.”An analysis cited in a recent Wall Street Journal article suggests that at least in some key areas like nanoscience, telecommunications, and energy storage, the United States is more dependent on Chinese research than China is on ours. Some researchers argue that an end to the science and technology agreement would come at a precarious time.Blumenthal is unswayed. “The agreement has outlived its usefulness,” he says. “It was put into place when the United States had an interest in China’s success as a great science and technology power, as a check on the Soviet Union. Now we know that China intends to wield its scientific and engineering prowess in geopolitically harmful ways.”The Chips and Science ActStirring up debates about China and basic research has been the Chips and Science Act of 2022, which emphasizes super-conductors and microchips but also has a much broader scope.Beyond computing, AI, data science, and cybersecurity, the act’s technology-focused areas, according to an analysis by the American Federation of Scientists, include robotics, automation, and advanced manufacturing; disaster prevention and mitigation; advanced communications and immersive technologies; biotechnology; advanced energy technology, storage, and industrial efficiency; and advanced materials science.The act addresses challenges involving the United States’ national security; manufacturing and industrial productivity; work-force development and skills gaps; climate change and sustainability; and inequitable access to education, opportunity, and services. It aims to enhance high-tech research to bolster national security and to bring microchip manufacturing back to America. (The acronym stands for Creating Helpful Incentives to Produce Semiconductors.)The Chips and Science Act was driven by increasingly tech-focused tensions with China, Russia, and other world powers, as well as microchip-production and supply-chain issues that arose during the pandemic and hobbled the auto, consumer-electronics, LED, and renewable-energy industries. The act aims to strengthen economic and national-security links in research, manufacturing, education, and work-force training, and to create tens of thousands of new tech-research and manufacturing jobs.The United States makes 12 percent of the world’s microchips, according to a recent report from McKinsey and Company consultants, down from 37 percent in the 1990s. American companies’ dependence on chips from abroad became all too clear during the pandemic. “Moreover,” the report states, “McKinsey research estimates that worldwide demand will keep growing, with semiconductors poised to become a $1-trillion industry by the end of the decade.”Over a decade, the act authorizes $250 billion in semiconductor manufacturing and broader research and development. As Jon Barnhart, director of research advisory services at the education consulting and technology company EAB, points out in a 2022 briefing and in a recent interview with The Chronicle, the act that leans on higher education to help carry out its goals without delivering much money directly to colleges. If colleges want a piece of it, he explains, they have to seek out funds allocated through various multiyear Chips-related streams — $81 billion at the National Science Foundation, $68 billion at the Department of Energy, $11 billion at the Department of Commerce, $10 billion at the National Institute of Standards and Technology, and, down the line, some support through NASA.Federal-funding streams reach colleges and their regions in myriad ways. For example, through Build Back Better funds from the Commerce Department’s Economic Development Administration, Wichita State University is part of a South Kansas Coalition that will receive $51 million for advanced manufacturing and research to strengthen American aerospace production by small and mid-sized firms.But the Chips and Science Act’s semiconductor emphasis has its own distinct impact. The act, which includes tax credits for chip manufacturers, has already spurred pledges of more than $200 billion in investments by TSMC, Intel, Samsung, Micron, and Texas Instruments. Those investments have ripple effects in academe beyond large research universities.For instance, to help TSMC and Intel meet their demand for workers, the Maricopa Community College District in Arizona is “currently working to train 4,000 to 6,000 semiconductor technicians from the local work force over the next few years,” the system’s chancellor, Steven R. Gonzales, wrote in the Arizona Capitol Times. Such positions, he explained, “are not just jobs — they are middle-class careers that offer pathways to promotion and upward mobility.”Of those students, 53 percent were first-generation college students, most of whom identified as students of color, and two of the cohorts were all women. “We’re proud to say,” Gonzales wrote, “that we are building a work force that increasingly represents the diversity of our state.”Despite such enthusiasm, some market analysts question the act’s likely overall impact.Yes, says a Goldman Sachs analysis, the demand for chips is expected to accelerate because “megatrends such as 5G, electric vehicles, artificial intelligence, and high-performance computing have positioned global semiconductor revenue growth on an even steeper trajectory than it has been on in the past several decades.”But, the critique says, building and running a chip-manufacturing plant costs 44 percent more in the United States than it does in Taiwan, due to higher capital expenditures, operating costs, and operational inefficiencies tied to cultural and other differences. Goldman Sachs forecasts that the act will increase chip production by less than 1 percent, creating a marginal “domestic pressure valve when future supply disruptions occur” but not reshaping the global industry.Such humbugging, however, doesn’t stop universities from charging full-speed ahead toward Chips opportunities, including $13 billion in research and training funds. “No sane university with a strong interest in microelectronics is sitting this out,” Jesús del Alamo, professor of electrical engineering at the Massachusetts Institute of Technology, told Science in 2022.Those funds fall under a consortium called the National Semiconductor Technology Center. Vying to be part of that consortium are “two large industry-academic-government partnerships,” Science reports: “The American Semiconductor Innovation Coalition (ASIC), spearheaded by IBM and New York’s Albany Nanotech Complex; and the Semiconductor Alliance, which features Intel, Micron, and the MITRE Corporation.” Affiliated with the former are MIT, the Georgia Institute of Technology, and Purdue University, and with the latter Stanford University and the University of California at Berkeley.In recent decades, as chip manufacturing increasingly migrated elsewhere, post-secondary students in the United States focused on chip design instead. The Chips and Science Act might help lure some of them back to manufacturing science and operations. That training could be an area where some Chips money can find its way to colleges other than powerhouse research universities.“Semiconductor research and manufacturing are highly specialized areas in which only a handful of established institutions operate because of high facilities costs,” says Martina Nieswandt, vice president for research at Chapman University in a written response to a Chronicle. So Chapman “focuses on how faculty’s work with undergraduate students is providing new avenues for technical training in the semiconductor industry.”Who Should Seek Chips Funding?Colleges eyeing Chips money should bear in mind that a quarter of the act’s authorized support is being channeled through the NSF’s new Directorate for Technology, Innovation, and Partnerships. That suggests an emphasis on applied and developmental research, EAB’s Barnhart says. If that’s what your faculty researchers are already doing great. If not, might be a mismatch.Moreover, as Chapman’s Nieswandt noted, most R&D related to microchips has a hefty initial price of admission. Some Chips money will find its way to institutions beyond the power players through NSF Regional Innovation Engines and technology and innovation hubs that help colleges collaborate with industry and regional governments.How far the spokes from those hubs will reach remains to be seen, when semiconductor-fabrication plants cost around $15 billion to $20 billion, and semiconductor-related research facilities, which require advanced cleanrooms, can cost up to $1,000 a square foot.While new Chips funding will offset some of these costs,” Barnhart has written, “the federal agencies continue to show a preference toward institutions with established facilities and track records of success when it comes to funding big-ticket, politically involved projects.”The biggest asterisk next to Chips and Science funding is that some of it may never happen. That’s always true of authorization bills when they get sliced, diced, and wrangled over in the appropriations process. But it may be even more true in this case because of discord in Washington over the size of the federal deficit — $2 trillion in 2023, double that of 2022 — thanks in large part to high-interest rates and low tax receipts.But even if funding doesn’t reach the bill’s goals, “it is important to set goals,” says Deborah Altenburg of the Association of Public and Land-grant Universities. “For our economic competitiveness and for the good of the country,” says Altenburg, APLU’s associate vice president for research policy and governmental affairs, “we want to spend more in these areas.”Barbara R. Snyder, president of the Association of American Universities and former president of Case Western University, applauds the way the act balances chip technology and the wider multidisciplinary-research agenda. Strategically, she says, “it’s very important to be looking around the corner.”Return to Table of ContentsSection 3: Priorities and PrestigeBalloons! Placards! Cheerleaders! Did the University of Louisiana at Lafayette just win a big game?Actually, the university’s 2022 celebration featuring all that revelry marked a different kind of occasion: It had been designated an R1 — or Research 1 — university, joining higher education’s top tier of research institutions as defined by the Carnegie Classification of Institutions of Higher Education.Yay!Wait … what?The Carnegie Classification system isn’t on the radar of many students, parents, and alumni, and even faculty, staff, and leadership may be hazy about it. But vice presidents of research watch it the way stockbrokers watch the Dow, because the classification affects a college’s reputation, position in high-profile rankings, ability to get grants, and power to recruit top-notch faculty members and graduate students. The classification is so important to college leaders, in fact, that research and higher-education experts fear institutions might sometimes skew their priorities to reach R1.Let’s look at the system’s evolution and its plans for the latest in a long line of modifications.How Did We Get Here?If you’re confused about the structure and nuances of the highly influential Carnegie Classification system, don’t feel too bad. Until recently, so was the guy who now runs it.Mushtaq Gunja, senior vice president at the American Council on Education, took over the system’s executive directorship in 2022 from Victor M.H. Borden, who had been at the helm since 2014. To prepare for that transition, Gunja, who is also an adjunct law professor and former assistant law dean at Georgetown University, huddled for many hours with Borden, a professor of higher education and student affairs at Indiana University at Bloomington.Before Gunja started to look closely at it, he, like many people inside and outside academe, thought the classification system was mostly about the size of an institution’s overall grants and research spending. Those are certainly part of the equation, but it’s much more complicated, as he would quickly learn.The system remains under the auspices of the Carnegie Foundation for the Advancement of Teaching, but just as Indiana University was its prior administrator, ACE is now. Gunja’s mandate from both ACE and the Carnegie Foundation, which has recently stepped up its involvement, is not just to run the process but to reimagine it. And, he says, changes are coming.“I would like to see a simpler, fairer, more equitable research methodology,” he says.Gunja, who was also a federal prosecutor and, during the Obama administration, a staffer in the Department of Education and the White House, is taking a consultative approach, working with 19 college leaders on an institutional roundtable and 16 experts on a technical-review panel, while also meeting with other high-level administrators at colleges, federal agencies, and elsewhere.The seven “basic” classifications are derived from federal data and describe an institution’s size, degree levels, and other fundamental characteristics. Those basic classifications are doctoral, master’s, baccalaureate, baccalaureate and associate, associate, tribal, and special focus.Those divide further into 33 classifications, including the R1 and R2 classifications for doctoral institutions. Institutions can also apply for “elective” classifications that showcase their diverse visions and strengths.Upcoming changes to the system should make the process more transparent and highlight features at a range of institutions, Gunja says. Carnegie announced some of those changes in November 2023, with further announcements and discussion to come.America’s social and political environment is tense, and skeptics question colleges’ value and contributions. New and future elective categories reflect important aspects of what many colleges already do. Higher education as a whole, and the Carnegie Foundation and ACE as two of its highly visible representatives, want to make that value and those contributions clearer, Gunja explains to groups with whom he meets.An elective community-engagement classification that began in 2006, for example, has picked up a lot of momentum and now includes about 360 colleges. Applications are being reviewed for the recently introduced Leadership for Public Purpose, which is a second elective classification. Institutions applying for that need to explain how they develop leadership, ethical judgment, and critical thinking “in pursuit of collective public goods like justice, equity, diversity, and liberty.”ACE will add one or more further classifications soon. A “universal” classification for every institution will reflect how well it advances social and economic mobility. Gunja and his colleagues don’t have a finalized approach on that yet, he says. They have been gathering feedback from the technical review panel on whether and how best to weigh, among other things, students’ persistence and completion; graduates’ earnings and debt repayment; access for students in various economic and racial groups; and geographical factors derived from census and other data.In the running for additional elective classifications are environmental sustainability, programs for learners connected with the military, indigenous-serving institutions, and programs for students who are incarcerated or otherwise affected by the justice system.ACE expects to present its updated classifications in late 2024 or early 2025.R1, R2, and Beyond“Research is amazing. I love our research institutions,” says Gunja, who follows vaccine development, new carbon-capture technologies, and planning for Mars expeditions with great appreciation. “It’s just that current methodology really recognizes a very particular type of research.”What began in 1973 as a descriptive matrix to help legislators and the public better understand and navigate America’s wide range of colleges has transformed, over the decades, from descriptive to normative. The resulting competitiveness around the classifications is understandable, Gunja says, but not always helpful, because it can distort institutional priorities.Despite the broad sweep of classifications, colleges and those who scrutinize them have often fixated primarily on the 146 R1 institutions and the 133 R2s, some of which strive mightily to become R1s. That’s because those classifications affect not just rankings and grants, but planning and benchmarking, faculty pay and recruitment, eligibility for other institutional memberships, and even criteria in athletic conferences..Classification terminology, over the years, has shifted this way and that, the R1 and R2 nomenclature officially disappearing and then unofficially returning. For five years, R1 became “research extensive” and R2 “research intensive.” Then, beginning in 2005, “very high” and “high” research activity defined the top tiers. But R1 and R2 seem to be what almost everyone still calls them.R1 and R2 distinctions were fairly simple in 1973, highlighting the research muscle of the top 50 R1 institutions, largely in terms of their federal funding levels, and the 50 institutions just below those in R2.Those formulas have become much more complicated in the half-century since. Key requirements for R2 now include awarding at least 20 research doctoral degrees, spending at least $5 million on research, and scoring high in a Research Activity Index calculation that weighs — across STEM, the social sciences, and the humanities — seven aggregate measures of research volume and three measures of research intensity. That intensity (sometimes referred to as density) boils down to research volume across the disciplines divided by the number of assistant, associate, and full professors.R1 institutions score even higher on that index formula, combining research volume and intensity across disciplines. How much higher is suggested by median research spending of $108 million for R1s and R2s combined in the 2020 fiscal year, compared with a median of $293 million for R1s alone.The revamp announced in November 2023 includes a significant shift in R1 and R2 criteria that will take effect with the 2025 classification. R1s will now need to spend $50 million on research and award 70 doctorates annually. R2s will still need to spend $5 million on research and award 20 doctorates. A new category, called Research Colleges and Universities, will include institutions that offer few or no doctoral programs but that spend at least $2.5 million on research.Under the new rules, the complicated research-intensity formula will be simplified to that number of new doctoral degrees awarded annually. And — in a departure from current criteria — it won’t matter whether the doctoral degrees and research spending are concentrated in a narrow range of disciplines or across the disciplines. That cross-discipline factor has been a benefit to some institutions and a stumbling block for others, so the new rules will affect strategies to get to R1. An early Chronicle analysis found that if the new rules had been in effect in the fall of 2023, the R1 category would have included 168 institutions rather than 146. And some current R1 universities would have lost that status, according to the analysis.The changes are meant, in part, to turn down the competitive heat — “unwarranted competition,” ACE calls it — and put the focus back on higher education’s rich institutional diversity, the roughly 3,700 or so American institutions that aren’t currently R1 or R2.Those include community colleges, regional state universities, teaching-intensive liberal-arts colleges, HBCUs, tribal colleges, technical colleges, online colleges, business colleges, professional and vocational schools, small rural colleges, institutions of faith or theology, arts schools, and so on. They all have their own missions, niches, histories, and challenges. Many are far from the orbits of big science, team research, high tech, or illustrious, endowment-fueled humanities projects — yet are still very serious about teaching students research skills and values. Shouldn’t the classification system, as initially intended, better illuminate all of those, too?College administrators are sympathetic to the reasoning behind the reforms and the new attention to access and social mobility, public leadership, and environmental concerns. Some wholeheartedly embrace what they’ve heard about contemplated changes.But others worry a bit. Will the changes distract from the research excellence they have worked so hard to foster?More research funding, higher-caliber research faculty, laboratories with better graduate-student and postdoctoral staffing, up-to-date equipment — those all help a college get to R1. And when it reaches R1, success can court further success and open up new opportunities. Research is expensive, and managing big grants productively proves that you can handle even bigger ones. Colleges assume that federal agencies and big foundations, even if they don’t explicitly say so, are reassured by top research credentials. After all, funding agencies want to make sure taxpayer and philanthropic funds are well spent.So a lot of planning and data collection have been designed around current criteria, and some administrators fear that a classification system with more components, even if some are elective, will muddle or dilute the R1 or R2 distinctions for which they’ve labored. Then again, once institutions reorient their record-keeping, the new formulas might allow colleges to highlight longstanding publicly minded efforts they haven’t had sufficient chance to showcase.At any rate, the Carnegie Classification system has been updated nine times since its founding, and college deans and vice presidents say they’ll roll with what comes.When R1 Is in SightTo be clear, not every college is obsessed with climbing the Carnegie Classification system.“I don’t think it’s top of mind, necessarily,” says Jessica Nicklin, associate provost for faculty affairs and research at the University of Hartford, classified as a doctoral university with moderate research activity.Hartford has federal grants and prides itself on specialized programs like its master’s degree in prosthetics and orthotics, one of 13 such programs in the United States, and its doctorate in physical therapy, which boasts 100-percent employment within 12 months of graduation.Nicklin is a professor of psychology whose research interests include workplace motivation and positive psychology. The university enrolls about 5,740 students, more than two-thirds of whom are undergraduates, and she and her colleagues try to motivate their students by bringing them along for the research journey.Sharing enthusiasm for research gets undergraduates, particularly students from marginalized groups, thinking proactively about honors tracks and thesis projects, says Nicklin. And coordinating with groups like a Hartford chapter of the Society of Women Engineers helps support women in fields in which they are underrepresented. But obsessively calculating where Hartford stands in the Carnegie Classification system wouldn’t particularly bolster those efforts.When R1 does beckon, it is often because institutions know they’re close enough to have a real shot at it, and they have growing research momentum already. Then the Carnegie research top tier can become a helpful framework for strategic planning and a way to fire up the troops.The University of Montana, which reached R1 status in 2022, had been rapidly expanding its research efforts over a decade. The university spread the word among its faculty about major federal grants, which inspired more grant applications in a constructive cycle, says Scott Whittenburg, the university’s vice president for research and creative scholarship. “People could see there really was something going on,” he says, sparking a “positive feedback loop.”Research spending more than doubled from roughly $55 million in 2014 to about $126 million in 2022, he says. On the way, around the $100-million mark, Montana realized it “probably could become an R1.”Carnegie metrics’ comprehensiveness across STEM, social sciences, and humanities fueled faculty enthusiasm. The university doesn’t have an engineering or medical school but is nonetheless strong in social sciences, pharmacy and other health sciences, forestry, ecology, chemistry, defense research, human-performance studies, natural-disaster prevention, smart sensors, and other areas, Whittenburg says. Recent faculty hires enhance the university’s standing in AI, software engineering, and machine learning.The University of Denver also reached R1 in 2022. It had been strong in military and space research during World War II and afterward, says Corinne S. Lengsfeld, senior vice provost for research and graduate education. But by the 1980s its research program had lost some steam.A more secure financial footing got the research wheels rolling faster about 20 years ago, and by 2013, with roughly $20 million in federal grants, leadership and the faculty alike were “shocked to see how really highly productive the university is.” That’s when they started more systematically bench-marking research growth, which in recent years has reached about 16 percent annually. “It becomes a faculty-led phenomenon,” says Lengsfeld. By 2023, research spending had grown to about $49 million.“We monitor the money, but we really look for how we affect the city, state, nation, and the world,” she says. “A private institution, dedicated to the public good,” is part of the university’s official identity and is practically a mantra for faculty and staff. Research priorities align with that, says Lengsfeld.The university has a relatively small campus with no medical school, and it isn’t known for STEM fields. As with Montana, though, the comprehensive cross-disciplinary Carnegie measures were a boon because of Denver’s strengths in the social and behavioral sciences. It doesn’t have a huge faculty, but out of 801 faculty members, 300 are engaged in research. Even once R1 was in its sights, the university didn’t tweak tenure and promotion standards to advance faster in the Carnegie metrics. “We did it,” Lengsfeld says, “just being who we were and who we wanted to be.”Lengsfeld is not sure students pay much attention to the R1 distinction, but, she says, “faculty applications went through the roof.” Since 2012, Baylor University’s strategic plans have included increasing its research base and enhancing the university’s national reputation, although R1 wasn’t mentioned specifically. Linda A. Livingstone became president in 2017, and the 2018 strategic plan aimed, she says, to make Baylor “the pre-eminent Christian university in the country.”Larry Lyon, who is vice provost for faculty recruitment and hiring as well as dean of the graduate school and a professor of sociology, had been tracking Baylor’s research growth by measuring it against Carnegie metrics. “We actually could attain R1 status,” Lyon, Livingstone, and their colleagues determined, “and we began to talk about R1 as one of the measures we could aspire to,” Livingstone recalls.They thought it would take five to seven years, but they got there in three. Baylor had a strong reputation in the social sciences and the liberal arts, Livingstone says, and investments and focus on STEM fields, particularly computer science, helped advance those areas. Baylor’s research spending for the 2023 fiscal year was $84 million, of which $64 million was for science and engineering fields, and $18 million for other disciplines.“We didn’t do R1 for the sake of R1,” Livingstone says, but as an empirical measure of the progress Baylor sought in its Christian mission to bring solutions to world problems.AAU’s Clout and MystiqueCarnegie classifications might strain your brain, but at least there are formulas behind them. The Association of American Universities, on the other hand, is a club, and while there are somewhat-murky criteria, the process is mysterious even to some colleges that have been selected for membership.“We were as surprised as anyone,” says John McGreevy, provost at the University of Notre Dame, which was one of six universities voted into the AAU in May 2023 for a total of 71 members.It had been a longtime goal, and Notre Dame had been amplifying its research presence in hopes of someday joining, but there’s no application process, says McGreevy, a historian who joined the faculty in the faculty in 1997 and rose through the deanship of the College of Arts and Letters.Membership is gratifying for all the faculty and staff members who have worked so hard, and networking with peer groups about the challenges inherent in running large research institutions can only be a boon, he says. The Roman Catholic institution, he says, might also be able to bring a different perspective to the AAU as a religiously affiliated university. (Brandeis University, also in the AAU, is formally non-sectarian, although it was founded in the 1940s by the American Jewish community.)As we’ve seen, standards to reach Carnegie R1 are stringent, and all AAU members are R1 colleges, but what makes them ostensibly the most elite of that impressive R1 group?In the AAU criteria, a few points stand out:“faculty awards, fellowships, and memberships,” including — starting in 2023 — membership in the national academiescitations of researchscholarly book publication, “especially in the fields of arts, humanities, and social sciences”AAU’s “informational metrics” include:enrollment of students receiving Pell grantsgraduation rates of undergraduatesgraduation rates of Pell recipientsthe gap in graduation rates between Pell recipients and other studentsConsidering these criteria, what emerges is a blend of top-of-their-fields scholarship and an intense focus on student success — not just what colleges are doing with benign intent, but what they’re doing effectively.With its growing size and diversity, AAU’s impact is often most evident on Capitol Hill, higher-education experts say. AAU’s “primary purpose,” listed as its first principle, “should be to provide a forum for the development and implementation of institutional and national policies promoting strong programs of academic research and scholarship and undergraduate, graduate, and professional education.”Note the word “national” there. AAU’s president, Barbara R. Snyder, says AAU would like to see the Pell Grant doubled, and she cites a white paper the association wrote with Business Roundtable on creating more-streamlined, flexible, and welcoming immigration policies for international STEM students.AAU’s role as a key national voice for academic research provides its greatest value, says Jon Barnhart, director of research and advisory services at the educational and technology company EAB.Higher education “probably couldn’t have gotten some of these bigger research funding packages through Congress without AAU’s strategically marshaled support,” he says. He compares AAU’s clout to Canada’s U15 group or Britain’s Russell Group.Even more influential, he says, is when AAU works in conjunction with groups like the Association of Public and Land-Grant Universities, the Council on Government Relations, the University-Industry Demonstration Partnership, the American Association for the Advancement of Science, and the Federal Demonstration Partnership.Together, these groups “carry a tremendous weight in D.C.”Nonetheless, Barnhart says he doesn’t try to guide client colleges on how best to angle for AAU membership because its membership practices are “opaque.”‘Elite, but Not Elitist'Snyder, AAU’s president, highlights the organization’s increasing diversity, with eight Hispanic-serving institutions, and colleges that are public and private, large and small.Membership trends do emerge, observers note — having a research-heavy medical center sometimes seems to help. Shalin Jyotishi, senior adviser for education, labor, and the future of work at New America, a Washington think tank, wonders if member institutions’ geography might factor into the equation in terms of congressional representation.Regarding membership, Snyder simply points to the official membership criteria and underlines that membership votes are carried out by AAU’s Board of Directors, who are, in turn, elected by AAU member presidents and chancellors. While there is no cap on the AAU membership number, she says, “we want to be a size where we can talk to each other in a constructive way.”Holden Thorp, editor in chief of the Science family of journals, former chancellor of the University of North Carolina at Chapel, and former provost at Washington University in St. Louis, applauds AAU’s recent expansion and growing institutional diversity. He worries, though, that it will make an even wider group of colleges contort themselves to try to get in, as they sometimes do for Carnegie’s R1 classification.Institutions that think they are contenders for AAU will, Thorp fears, invest in doctoral programs in which they don’t have a logical advantage. Not every college needs to excel at neuroscience, for instance, if their research strengths are in metabolic diseases or public health, he says.Ishwar K. Puri, senior vice president for research and innovation at the University of Southern California, says AAU is “elite, but not elitist.” Its discussions overlap with those at APLU and other associations, he says. As part of an AAU group of senior research officers, he finds that institutions with large research enterprises share particular challenges. Among those, he says, are the dangers of foreign influence, facility security and cybersecurity, immigration and other policies around international researchers, and truly daunting layers of regulatory compliance protocols.And as with R1, the satisfaction of being admitted to AAU is tempered by the fear of losing the distinction. It’s right there in the fourth principle in AAU’s Membership Policies:“All current members are subject to ongoing assessment by the Membership Committee. In those instances in which there appears to be a significant and sustained disparity between the mission or accomplishments of a member institution and that of other members of the association, the Membership Committee will make the appropriate recommendations to the board about continued membership.”Over the decades, Clark, Catholic, Syracuse, and Iowa State Universities surrendered their memberships, a move one expert described to Inside Higher Ed as a pre-emptive effort to “save face.” The University of Nebraska was voted out — and is now vying to get back in.One glaring absence in Carnegie’s R1 classification and AAU is a historically Black university. Howard University may, before long, accrue the numbers to cross back into R1 for the first time since a Carnegie shift in criteria knocked it out of that category in 2005. In fact, The Chronicle’s fall 2023 calculations show that Howard had already met the revised R1 criteria. In Section Five of this report, we’ll look at one project that is helping Howard raise its research profile.Return to Table of ContentsSection 4: Partnership Models, Old and NewWe saw in Section One that policymakers debate the best balance between basic and applied research, and whether the diminished proportion of funding for basic is problematic.Here, we will look at how effectively applied research is being disseminated into society through collaborations with industry and government. We’ll see a mix of old and new programs and paradigms.First up, a huge old-school model that has yielded generations of cutting-edge results.As mentioned in Section One, the Johns Hopkins University invariably tops the list of universities’ research and development spending — by a lot. In 2021, it spent $3.18 billion. Second on the list is the University of California at San Francisco, at $1.71 billion.Why is Hopkins in a spending orbit of its own? The university’s Applied Physics Laboratory, or APL, with $1.91 billion in R&D spending for the 2020 fiscal year. The lab began in 1942, as part of the federal Office of Scientific Research and Development, to devise technologies for the war effort. Its first home was a used-car garage in Silver Spring, Md. Today, at its 461-acre campus in Laurel, Md., the not-for-profit lab employs more than 8,600 people and is the largest of America’s University Affiliated Research Centers, operated under the Department of Defense.During World War II, the Hopkins lab developed a radio-proximity fuse that allowed U.S. ships in the Pacific to more effectively defend themselves against aerial attacks; it was later used by ground troops in Europe. With the fuse, an explosive device detonates in proximity to its target and does not require a direct hit.In the eight decades since the lab’s thousands of contributions to national security and space-exploration projects have included its work on the first satellite-based global-navigation system, the Satrack strategic missile-tracking system, the onboard-guidance system for the Tomahawk long-range cruise missile, sea-based ballistic-missile defense, and planetary defense technologies to use against asteroids on a collision path with Earth. Other projects include neurally controlled prosthetics, cybersecurity, and global health surveillance.Less old school than the Hopkins lab but dating back, in its current form, to the Bayh-Dole Act of 1980 is tech transfer. That legislation allowed universities to profit from federally funded research. In the 43 years since, tech transfer has yielded more than 117,000 U.S. patents and 15,000-plus start-ups, contributing $1 trillion to U.S. gross domestic product and supporting six and a half million jobs, according to AUTM, the tech-transfer association.Despite those numbers, and some oft-cited blazing tech-transfer success stories like Google and Gatorade, some critics think the model has not lived up to expectations. One expert quoted in a recent Chronicle Trends Snapshot said tech transfer’s impact on science has been “negligible, if not negative,” accounting for less than 4 percent of patents and 1 percent of start-ups nationally, with most university tech-transfer operations losing money.Venture-development firms, in some instances, help bridge the gap between initial innovation and seed money toward commercial development. But overall, the mind-sets, interests, skills, and temperaments of researchers have proved to be significantly different from those of start-up business executives, and bridging those differences remains tricky.Does that mean universities’ contribution to the market realm has been a failure? No, say university research leaders. It means that the translation of university discoveries into commercial products is only one way to measure impact, and probably not the most important one.“Part of me thinks we’ve really blown the metrics entirely,” says Corinne S. Lengsfeld, senior vice provost for research and graduate education at the University of Denver, where she is also a professor of mechanical and materials engineering.The most obvious contribution of academe to business, hiding in plain sight, she says, are its Ph.D.s. The vast majority go directly into industry with state-of-the-art knowledge and skills. Moreover, research professors creating innovations to be handed off to industry via patents isn’t the primary way scholars work with business, Lengsfeld says. About 20 percent of her portfolio is coordinating contracts between businesses and the university’s research faculty. Companies “absolutely see the value in this relationship,” she says, essentially using universities as their research arm “because we’re pretty cost-effective and they want an employee pipeline.”That said, she wouldn’t write off the tech-transfer model, either. It might have taken a couple generations for it to hit the right cohorts. Junior faculty, grad students, even undergraduates and high-school students “grew up with that start-up model.”“Do we do it perfectly? No.” But, she says, “you know, universities are pretty motivated. We’re in a time when we need another revenue stream. I think what you’re going to see is universities experimenting in ways that they haven’t.”Northern Indiana Gets a LiftIn conversations about the intersection of academe, industry, and government, there’s a phrase that, while not eclipsing commercialization or tech transfer, is often mentioned alongside them. That phrase is knowledge exchange.It can include entrepreneurial and other business ventures but puts equal emphasis on worker training, regional development, and the sharing of up-to-date practical and theoretical know-how.Jon Barnhart, director of research advisory services at the educational consulting and technology company EAB, says the idea started gaining traction during the Great Recession. Government dollars shrank and venture capital was hard to come by. Universities needed new ways to usher their knowledge into the world. There was suddenly less talk of those elusive angel investors and more discussion of creating a whole innovation ecosystem, he explains.“It doesn’t just mean you put cool ideas in the market and hope someone picks them up,” he says. Knowledge exchange is focused, intentional, detailed, and it requires sustained attention.NSF Regional Innovation Engines and hubs, and the regional coalitions under the Commerce Department’s Economic Development Administration, are examples of government-generated knowledge exchange that include colleges and universities. But there are more homegrown organic instances too that aim to foster education, job training, regional development, and business growth.A case in point is the University of Notre Dame, which has been working with local businesses and governments in the South Bend and Elkhart, Ind., region to share knowledge through a program called LIFT.That stands for Labs for Industry Futures and Transformation, a regional project supported by a $42.4-million grant in 2019 from the Lilly Endowment. LIFT is a collaboration of civic, institutional, and industry partners in the region that is led by Notre Dame, the South Bend Elkhart Regional Partnership, and enFocus, a nonprofit that aims to attract professional talent to northern Indiana through fellowships and internships.LIFT’s goal is to enhance the region’s innovation ecosystem. It aims to increase regional productivity by introducing advanced technologies to local industry; boost regional resilience by spurring growth in emerging sectors like IT, data analytics, electric vehicles, and aerospace; and develop, attract, and retain talent in the region for existing and emerging industries.At the heart of Notre Dame’s participation is iNDustry Labs, which includes facilities and members of the university’s faculty and staff who act as engineers in residence, providing technical-advisory services to local firms. Those engineers in residence also link businesses with the university’s academic and commercialization resources.Scott Ford is the executive director of iNDustry Labs and Notre Dame’s associate vice president for economic development. The faculty director for iNDustry Labs is Thomas Fuja, a professor of electrical engineering at the university.Northern Indiana is all about manufacturing, they explain. When many factories moved abroad, unemployment shot up and the area experienced a brain drain. Manufacturing that has remained, especially in the recreational-vehicle industry, is modernizing, with a focus on automation, digitalization, and diversification. It’s in that light that Notre Dame and area businesses can exchange expertise, training, and work experience, hopefully increasing the region’s worker retention in the process.Among LIFT’s achievements from 2019 through 2023, its organizers say it:spurred $88 million in new facilities and equipment, and $116 million overall, with every one dollar in seed money yielding five in direct investmentsfunded 17 full-time positionslinked iNDustry Labs with 63 companies on 169 projectsplaced 178 students in internships with 23 companiesplaced 488 post-secondary students — 209 from Notre Dame, the rest from eight other partner colleges and universities — in experiential-learning courses with regional businessesengaged 35 enFocus fellowships with iNDustry Labs projectsThrough partner colleges and universities, LIFT has 1,856 credential and certification course enrollments in advanced technical skills completed or underway. LIFT programs have enrolled 165 students in new and expanded STEM degree programs.The tens of millions in investments from Indiana and then the Lilly Foundation are paying off, LIFT says. While it is still too early to draw firm conclusions, local industry partners have identified over $10 million in new investments in advanced technologies as a result of industry collaborations. Similarly, from 2019 to 2023 there has been a fourfold increase in the amount of industry-sponsored research at Notre Dame sourced from local companies. According to data from Lightcast, a labor-market analytics company, the number of STEM degree completions in the region’s nine colleges and universities has increased by 5.5 percent over the same period.From Forklifts to Futuristic FreewaysInnovation, then, sometimes comes jumbo-sized, as with the Hopkins Applied Physics Laboratory. Sometimes it’s motivated by patents and potential startups, as in the Bayh-Dole model. And, as in northern Indiana, it can be fostered through regional collaborations like those among Notre Dame and its neighbors.Innovation can also occur in a smaller, more direct way when catalyzed by a single company — a company like The Raymond Corporation, a century-old materials-handling firm in Greene, N.Y., that is a member of the Toyota Industries family of companies.Unless you buy forklifts or need advice on warehouse logistics, you may not have heard of Raymond. A number of university researchers have, though, because of Raymond’s University Research Program.The company has a history of innovation. Wooden pallets and the lift trucks that carry them? Raymond invented those, and they were patented on the same day in 1939. The company has been innovating ever since, with hundreds of configurations of dozens of lifts, and related equipment, automation, optimization, and consulting services. Raymond introduced microchips to its lift designs in 1986, and more recently alternating-current models, hydrogen fuel cells and lithium-ion batteries, plus a virtual-reality simulator to train lift-truck operators.But how, after a century, do you keep that spirit of innovation going? Raymond executives asked themselves that question around 2015. You go to major research universities, ask professors what they’re passionate about, and see whether it has links to material handling.As it happens, a lot of research does. Automation; autonomous (self-driving) vehicles; alternative, clean, and more-efficient energy solutions that work in indoor space; virtual training and remote operation; supply chains; untraditional retail environments and shipping — forklifts, weirdly enough, touch on many key research issues of our era.And so it is that, since 2016, over six funding cycles, 45 professors from 22 universities have worked with Raymond on 15 projects. By government-grant standards, the $4 million-plus that Raymond has awarded for the program is modest. Researchers like these grants, however, because they can keep any patents or other intellectual property that stems from their work. Raymond retains royalty-free license to use the technologies.The researchers also like that they have considerable leeway to adjust the course of their Raymond-sponsored project if it pulls them in some interesting new direction. And unlike most government-grant projects, Raymond’s involve helpful monthly check-ins with company personnel to keep ideas flowing and get feedback from industry specialists who know their equipment, literally, inside and out. With a federal grant, the researchers say, there’s often less flexibility, less brainstorming, and it’s usually more like, “See ya in a year. Good luck.”Steven Medwin worked full time for Raymond from 2001 to 2019, with leadership roles in test engineering, advanced research, systems engineering, new-business development, and energy solutions. In the last four years, he’s worked part time coordinating the research program as university ambassador.The problem with traditional tech transfer, says Medwin, who has degrees in mechanical engineering and business and also worked at DuPont and the Enser Corporation, is that businesses think in terms of near-term profits, while academics want years to “work on the coolest, most esoteric minutiae that hasn’t been done before — they survive in this small niche.”His role is to straddle those two worlds. “How do we get the professors to come out of their niche and say, ‘Hey, there’s a real-world problem we need you to solve,’” while getting corporations to be patient and give the researchers the freedom they need to follow their instincts?The participating professors are usually part of a research team, and the Raymond grant is for one among a cluster of the research team’s projects.Jennifer Pazour is a professor of industrial and systems engineering at Rensselaer Polytechnic Institute. She studies how mathematical models can improve logistics and supply chains. For The Raymond Corporation, Pazour’s lab, in collaboration with a lab in Germany, is studying how new equipment and facility designs can improve order fulfillment in retail stores.Imagine robots picking items off shelves, then shuttling dry, canned, and bottled goods to you after you’ve selected your bananas and avocados and just the right rack of lamb because you’re pickier about those and need to eyeball them yourself. At the walk-through exit, everything’s automatically rung up for you.Those are the kind of operational scenarios Pazour and her colleagues experiment with in detail. What kind of robotics, store designs, travel paths, shelf-stocking patterns, open and hidden storage spaces, and past-purchase technologies should they entail?Pazour’s work with the Raymond grant is not a consulting project but a research one, she says. “It’s aligned with the business model but it’s more forward thinking.” Because Raymond encourages researchers to publish, she says, it is “helping the whole industry.” And, she says, she likes that she got to propose what interests her instead of contorting her work to fit the precise conditions of a request for proposals, the way many industry research projects work.Forklift operators, especially since the pandemic, are in demand, as are efficient methods to train them. That’s what Sol Lim, an assistant professor of industrial and systems engineering at Virginia Tech, is studying with her grant from Raymond. She specializes in human-factors engineering and ergonomics (the study of how people interact with their work environments).Lim’s team observes experienced drivers operating forklift trucks using a Raymond-developed simulator, and measures their body movement, brain activity, eye movement, and driving performance in comparison with those of novice drivers.The team’s results are preliminary, but as you might expect, Lim says, the more experienced a driver is with a certain machine, the smaller the cognitive load because operation has become more routine. As you might not expect, though, more-experienced drivers move their bodies more, not less, because they try to gain more situational awareness. They rotate their bodies a fair amount to avoid hitting shelves and people.Next up, says Lim, are experiments with multimodal feedback, providing visual and haptic guidance to novice drivers — a vest, for instance, that vibrates on one side or the other of an operator’s torso if the operator is too close to an obstacle. The goal is to incorporate expert drivers’ skills into virtual-reality training to enhance novice drivers’ skills faster.Then again, some forklifts don’t have operators. They are autonomous and can work around the clock. Except they can’t quite work 24/7 because they have to stop for battery recharging.What if they didn’t have to stop? What if they didn’t have batteries? What if they were propelled by wireless electrical power through the floor?Khurram Afridi is an associate professor of electrical and computer engineering (ECE) and director of graduate studies for ECE at Cornell University. His Raymond grant, along with NSF and Department of Energy grants in the area of wireless power transfer, constitutes about a third of his lab’s research. His team consists of seven Ph.D.s, a postdoc, three master’s students, and four undergrads. Several people on that team are working on wireless power transfer.The wireless power Afridi thinks will someday work for large vehicles like forklifts is transmitted via electric fields by metallic plates, not via magnetic fields by coils like the wireless charging platform on which you place your cell phone. Two metallic plates in the flooring are mirrored by two in the base of the forklift. That’s with no battery. Or, with a small battery, the vehicle can get over areas where there are no plates until it reaches the next set of plates.There’s a long way to go, Afridi explains. Powering a 9,000-pound forklift is very different from powering a 180-gram phone. Scaling up the kilowatts, increasing the efficiency of the power transfer over an air gap between floor and vehicle, controlling the direction of the electric fields to keep a system safe for the forklift’s operator and others, the cost of making and installing metallic plates in warehouse flooring — those are just some of the challenges in scaling up the work he and his colleagues are conducting.Then again, until around 2015, people had written papers about why the electric-field approach couldn’t be done, he recalls. And now he’s doing it.And by the way, Afridi’s federal grant money? Some of it is to study similar wireless charging for cars on highways. Imagine: As your great-grandkids shuttle their families in autonomous vehicles from Ithaca to Washington, D.C., they might not even need a charging station. The Afridi Highway will charge their cars as they go.Return to Table of ContentsSection 5: 4 Red-Hot Research RealmsWe’ve looked at the big-picture research funding stream, at how universities strategize around Carnegie classifications and other markers of prestige, and at several kinds of partnerships that researchers and funders use to usher innovations beyond the campus gates.Now let’s visit some labs in four areas of particularly intense research:energy solutionsrobotics, biotech, and healthartificial intelligencequantum computingAlternative EnergyRenewable energy is on a roll.Global investment in clean energy was projected to reach $1.7 trillion in 2023, according to the International Energy Agency, with solar, for the first time, eclipsing oil production. Renewable energy is projected to surpass coal by 2025 as the largest source of global electricity, The New York Times explains. China is the top producer of wind and solar power; one-third of England’s power is generated by wind, solar, and hydropower; and in the United States, 23 percent of power is generated from renewable sources, up 10 percent from a decade ago. Meanwhile, the long-term potential for green hydrogen and nuclear fusion is drawing significant investment, The Wall Street Journal reports.But underlying the splashy, headline-making industrial-energy advances is slow, steady, meticulous, foundational research on energy storage, sources, and sustainability, and much of that is happening at universities.Let’s look at two examples.‘The Belly of the Beast’Fuel oxidation occurs when a fuel source reacts with oxygen, releasing electrical energy and byproducts. It’s hard to improve that process without understanding exactly how it happens. When studying alternative-energy systems, examining a reaction that occurs at up to 800 degrees Celsius (1,472 Fahrenheit) was, until recently, impossible. Inferences could be made from the compounds that went into a reaction and those that came out. But to increase efficiency, scientists needed more information about the chemical and material processes.Anyone who pays an energy bill or worries about the environment will understand why a detailed picture of fuel oxidation is important. For the Air Force, which consumes about two billion gallons of aviation fuel annually and is the largest energy user in the Department of Defense, that is particularly true. Incorporating energy-efficient alternatives might lower costs and environmental impact.As a postdoc at Wright-Patterson Air Force Base, in Ohio, from 2011 to 2013, Bryan C. Eigenbrodt studied details of fuel oxidation and has continued to do so at Villanova University, where he is an associate professor of chemistry.Eigenbrodt and his students examine fuel oxidation in alternative-energy systems using a technique called Raman spectroscopy, which identifies molecules by the way they vibrate. Eigenbrodt has custom-built an apparatus that combines the spectroscopy with electrochemistry to develop a full picture of the catalytic processes occurring in solid oxide fuel cells. He can observe molecular changes at high temperatures and in real time.“We looked into the belly of the beast and lived to tell the tale,” Eigenbrodt laughs.Eigenbrodt’s lab uses a method known as operando spectroscopy to simultaneously analyze what a material is and how good a catalyst it is — in this case, how well it triggers oxidation. The method can determine the effectiveness as catalysts of perovskites — a group of minerals with strong but porous crystalline structures. Already used for some solar fuel cells, perovskites have potential for wider application, Eigenbrodt says. A perovskite anode — a fuel cell’s negatively charged electrode — is a “very tailorable” material, he says. You can increase conduction to get more power by substituting chromium for iron, for instance, or you can increase the catalytic property by substituting nickel for molybdenum.Suppose scientists figure out how to cleanly use ethanol as a fuel source. Where, in the vast quantities needed, would it come from?Corn is a main source now, but that takes away from corn as a staple food source. An alternative: algae, which leads us to the Eigenbrodt Research Laboratory’s other main focus.You’ve heard of the “freshman 15” or, Eigenbrodt sighs, “the faculty 40”? It turns out that algae, too, gain weight — retain lipids — when they are stressed. Almost all fossil fuel comes from prehistoric algae, he explains, so why are we still drilling to create fuel instead of growing new fuel that is sustainable and can be used more cleanly? Why not farm algae near shorelines?It grows there anyway, so could we not harness that green gooey life force for the oil needed not just for fuel for transportation but for the polymers in our cellphones and for everyday plastics?If you deprive algae of nitrogen, it stops producing chlorophyll and instead makes more triglycerides — fats that can be converted into the fuels we are currently using and depend upon. The problem of deadly algae blooms could now become an energy solution.In both the fuel-cell operando spectroscopy and the algae experiments, Eigenbrodt says, “we are trying to get an unprecedented look, and get unprecedented data, so we can move fuel cells and sustainable biofuel feedstocks forward for the future.”Go With the FlowTwo ways to shift the electric grid from its dependence on fossil fuels are through solar and wind energy, but sometimes the sun doesn’t shine and the wind doesn’t blow. So you store the energy in batteries, but lithium-ion batteries require precious metals that are hard to come by and expensive.Another kind of battery might work better. In redox-flow batteries, chemicals combine with electrodes, creating ion transfer — an electrical charge. In theory, those batteries can store energy on a much larger scale than can lithium-ion batteries. Scaling up, however, requires redox-active organic molecules — so named because an electron is transferred in an oxidation-reduction reaction — that are soluble enough to store in a liquid solvent.“These molecules are excellent at holding an electric charge,” David P. Hickey, an assistant professor of chemical engineering and materials science at Michigan State University, recently explained for an MSU article, “except they don’t dissolve well in the desired battery solvents. The amount of energy that can be stored is directly related to the solubility of its redox-active molecules, so more soluble molecules means higher energy density.”Hickey and Tom Guarr, director of research and development at the Organic Energy Storage Laboratory at Michigan State’s Bioeconomy Institute and cofounder of Jolt Energy Storage Technologies, recently discovered a way to make the redox chemicals more soluble. They found that carbon-hydrogen groups are attracted to so-called pi electrons in certain hexagonal ring structures of organic (carbon-containing) molecules. That disrupts electrostatic forces and makes the redox chemical more soluble. The more such interactions the redox chemicals have, the more soluble they are and the more energy can be stored.Hickey’s lab group, which includes five doctoral students, a master’s student, and a half-dozen undergraduates, is working with Guarr’s team and the National Renewable Energy Laboratory to put their findings to large-scale industrial use. Toward the goal of storing large quantities of solar and wind energy more cheaply and effectively, the researchers plan to scale up production of the soluble redox compounds using 4,000-liter reactors at the Bioeconomy Institute.His team’s work, Hickey tells The Chronicle, comes down to “using molecular engineering to solve big problems.” And, he says, “there’s not really a bigger problem than climate change.”Robotics, Biotech, and HealthBig science increasingly involves little science — that is, big scientific innovation at small scale. In robotics, that might mean looking to fascinating little critters for inspiration. In biotech, it might mean manufacturing customized mRNA, measured in nanometers, for treatments specific to one patient or one rare disease. And in immunology, it might mean adapting a successful cell-transforming treatment for one condition to treatments for other illnesses.Let’s visit three labs where those developments are happening.Evolution as EngineerIt’s been 69 years since the inventor George Devol’s Unimation, the first robotics company, developed Unimate, a robot with a hydraulic arm to lift heavy loads. Today, robotics is a $115-billion industry expected to grow to $258 billion within the next five years, according to Mordor Intelligence, an advisory firm based in India.A Maruti Suzuki car plant in Sanand, India, employs some 5,000 robots, Mordor reports. Mitsubishi developed a robot work-task system in 2022 that can operate as quickly as people and is controlled by voice commands. With a doctor operating it remotely from a console, the da Vinci Surgical System is used worldwide in hundreds of thousands of prostate, cardiac, gynecological, and other surgeries.And yet robots, especially mobile ones, often move and react awkwardly. K5, a new, 400-pound surveillance robot patrolling the Times Square subway station in New York, was scornfully referred to by the head of a privacy and civil-rights group as “a trash can on wheels.”What would happen if, instead of building robots around tasks, we developed them to respond to their environments, borrowing from the greatest engineer of all: evolution? That would require a bottom-up approach instead of a top-down one. Through three decades of collaboration, that’s the strategy taken by Roger D. Quinn and Hillel J. Chiel of Case Western Reserve University.Quinn is a professor of engineering and director of the Center for Biologically Inspired Robotics Research. Chiel is a professor of biology, neurosciences, and biomedical engineering. They are coordinating an international network of about 80 researchers, including 15 principal investigators, in the United States, Germany, and England in an $8-million project backed by the National Science Foundation, with additional funds supporting the European researchers. For the project, called NeuroNex: Communication, Coordination, and Control in Neuromechanical Systems, the international team is examining how animals in groups of different sizes coordinate and control interactions with their environment.“What a piece of work is man,” wrote Shakespeare — but hey, don’t knock sea slugs, fruit flies, and rats, cats, and other small vertebrates.“Biology is kind of an alien technology, and because we are biological organisms, we don’t recognize this,” Chiel explains. Each species has a neuromechanical system of its own, adds Quinn.In the late 1980s, Quinn and Chiel worked with Randall Beer, now a professor of computer science and informatics at Indiana University at Bloomington, to build a robotic insect, complete with an artificial nervous system. Since then, through various scientific meetings, they’ve established another organic network — one of colleagues abroad with similar interests. Now, for NeuroNex, these clusters of scientists and engineers will work on their own projects while communicating regularly to share insights.By learning from evolution, the researchers hope to engineer robots so that they operate from an internal neuromechanical logic rather than an imposed training logic. The group also hopes to derive computational models based on how various animals act naturally.The NSF is currently funding three additional NeuroNex networks. One network studies how organisms sense, process, and react to smells. Another examines how the strength of nerve synapses (the passing of a signal from one nerve cell to the next) relates to nerve structures and activity. And a third researches how the brain’s neocortex — thought to be crucial to sensation, action, cognition, and consciousness — factors into primates’ retaining of information for brief periods, which is key to abstract reasoning.Robots and their inner workings are becoming more sophisticated, with soft and squishy materials sometimes derived from 3D-printed substrates interlaced with lab-grown muscle sourced from mouse, insect, or sea-slug cells. Yet Chiel says the fundamental goal of his group’s NeuroNex project isn’t to create snazzy new biorobotic mechanisms — though that will likely happen — but to further “an ongoing conversation between robotics and biology.”Gene GenieIf roboticists take inspiration from biology, bioengineers are ingeniously manipulating the nanomechanics of genes.Philip Santangelo’s laboratory at Emory University is the first recipient of a grant from the National Institutes of Health’s Advanced Research Projects Agency for Health. ARPA-H, as it’s known, was established in 2022 to channel major resources in a push to cure stubborn medical challenges including cancers, life-threatening infections, and autoimmune diseases like lupus.Santangelo, a professor of bioengineering at Emory University and the Georgia Institute of Technology, will work with colleagues at Yale University, the University of Georgia, and a German company called Transimmune under the three-year, $24.8-million grant.He aims to develop a new class of mRNA-based drugs that can be personalized. With close monitoring of results as treatment proceeds, the therapeutics would turn genes on and off in individual immune cells without altering the recipient’s overall DNA. He also plans to create antigens and gene modulators to improve general vaccines and treatments.Santangelo and his Emory team of about 15 colleagues use advanced imaging to observe problems, down to the level of individual molecules, in mRNA regulation of inflammation, viral infections, and the growth of cancers. The imaging can help spot cancers in early stages, and it provides models for the synthesis of mRNA to treat or prevent disease, alter cells, and spur reactions in and on cells.“Not everything we do is going to work out,” says Santangelo, but if a few drugs prove successful, it could have a significant impact.Creating the drugs is only part of the puzzle, he says. Then comes affordable access and distribution.“As winners shake out, there will be tough discussions about moving them into the clinic,” he says. “I’m fairly familiar with that process, and I know it’s not going to be trivial. … There are always challenges with new technology.”Following a Hunch in Treating HIVThe smartest innovations sometimes stem from adapting solutions for old problems to new problems. Why reinvent the wheel when it’s wiser to adapt the wheel to new terrain?Christina Gavegnano was a doctoral student at Emory when she heard a lecture about an arthritis treatment using JAK inhibitors — molecules that bind to the Janus kinase enzyme, inhibiting a signal that would lead to inflammation.Working at the intersection of immunology and pharmacology, she wondered if JAK inhibitors might be useful in treating HIV. She still has the November 10, 2010, email, titled “HIV Eradication Idea,” that she sent her boss.Now directing the Gavegnano Drug Discovery Program at Emory, when she looks at that email, “it’s the weirdest feeling because we’ve come so far.”That 2010 hunch became Gavegnano’s dissertation, and her lab was the first to demonstrate the anti-HIV effectiveness of two JAK inhibitors called ruxolitinib and baricitinib. Monica Reece, a doctoral candidate in Gavegnano’s group, recently presented findings at the International AIDS Society Conference in Brisbane, Australia, showing that JAK inhibitors reverse immune dysfunction and reduce the HIV reservoir in treated patients.Baricitinib has also received FDA approval for the treatment of hospitalized Covid-19 patients, for whom it confers a 46-percent reduction in mortality, greater than that of any other treatment. Gavegnano says there may be uses for JAK inhibitors against other viruses and even drug-resistant clinical depression, which in some cases may be driven by inflammation in the brain.Gavegnano has 19 undergraduates in her working group of 27. They take part in lab work, drug trials, patents, and publications. “They have a front seat to all of this. They enjoy it. They become part of it.” She observes them, semester by semester, becoming more confident. “I find a lot of personal satisfaction watching that,” she says.But the experiences she shares aren’t all in pharmacology and immunology. Some are about persistence.“Up and down is the story of this journey, let me tell you,” she says, echoing many researchers’ tales of discouragement when they followed scientific instincts that went against the grain. “It was years trying to fight to get anyone to fund this, trying to get anyone to pay attention.”When Gavegnano earned her Ph.D. in 2012, the reigning theory was that HIV should be treated by allowing inflammation, then attacking the HIV reservoir with antivirals. It was akin to flushing out the enemy into plain sight so that you could attack it.Gavegnano recalls a meeting around that time when she was scheduled to present her work on an upper floor of a conference center. Climbing those steps to the small presentation room, she felt like a salmon swimming upstream, she recalls, because everyone else was storming down to a huge auditorium below. That’s where researchers were discussing a just-announced breakthrough that they thought would cure HIV. Only weeks later, that approach, involving two patients in Boston, was deemed unsuccessful.About 14 people, meanwhile, had shown up for Gavegnano’s talk. But she told herself and her colleagues, “We’re slow and steady. We’re going to keep going, and we’re not a flash in the pan.” Repurposing arthritis drugs for antiviral uses “doesn’t sound fancy,” she says. But it has saved lives.“Now we’re in a beautiful place,” she says,” with “scientific credence to move forward with this drug.”Artificial IntelligenceForgive AI, for it has sinned.It steals jobs and intellectual property. It helps propagandists lie and cybercrooks hack. It’s politically biased. When it trains robots, they’re racist and sexist. It encourages anorexic behavior. It makes up nasty stuff about you and posts it on the web. It hallucinates and gets progressively dumber. Terrorists will use it to disrupt and menace society. It will cause pedagogical chaos. When it inevitably nannies your children, it risks warping their brain development. It’s too unreliable, tactless, and insensitive to take part in health care. It’s developing consciousness and superhuman abilities but without moral constraints. It was hyped in the stock market, and its bubble is already bursting. It’s unsafe for corporate use. It requires unsustainable resources. And it’s not smart enough to manage your investment portfolio.No, wait: AI is vital for drug discovery and helped researchers identify a possible treatment for a metabolic disease that kills children. It helps predict landslides. It plays back a Pink Floyd song from brain scans and might help those who cannot speak communicate. It makes big construction projects more efficient. And it’s going to make your wonky children rich.A journey through most days’ AI headlines causes whiplash. But “man is neither angel nor beast,” wrote the French 17th-century mathematician, physicist, and philosopher Blaise Pascal. Nor is AI. It is an extraordinarily powerful tool as good or as bad as those who use it.On the whole, academic researchers and the universities that employ them are embracing AI with tremendous excitement and — for institutions that can afford it — major investments.The University of Florida in recent years has added 100 new AI-focused faculty members to the several hundred already working there, many in conjunction with its recently completed HiPerGator supercomputer.“We’re doing AI in medicine, AI in drugs, AI in agriculture, AI in business,” Joseph Glover, the university’s provost and senior vice president for academic affairs, said on a podcast in 2021. “The College of Business just made AI a required introductory course for its entering freshmen. ... We believe that this is going to be a transformational initiative for the University of Florida. We think that this is where higher education is going to inevitably go.”In August 2023, the Johns Hopkins University announced it was opening a new data-science institute in a new custom-built facility, with 80 new affiliated faculty members at the Whiting School of Engineering and 30 new Bloomberg distinguished professors.“Data and artificial intelligence are shaping new horizons of academic research and critical inquiry, with profound implications for fields and disciplines across nearly every facet of Johns Hopkins,” said Ronald J. Daniels, the university’s president.Those are just two examples, with the prevailing sentiment captured by Ishwar K. Puri, senior vice president for research and innovation at the University of Southern California, in a March 2023 announcement of its new Center for Generative AI and Society. “The genie is out of the bottle, and we’re not going back,” he said. “Our students are going to use this technology, and we need to teach them how to use it responsibly. We want to harness the power of AI for the public good.”Ravi V. Bellamkonda, provost and executive vice president for academic affairs at Emory University, thinks the primary role of universities in AI is less in the generation of new algorithms — it’s tough to compare with the Googles and Amazons of the world in that arena — than in applying AI to the liberal arts and other fields of human endeavor. Prize artificial-intelligence researchers in academe, he says, are versed in AI and other domains — like medicine, law, social justice, and literature.In the early days of biomedical engineering, there was a steep learning curve, says Bellamkonda, a bioengineer and neuroscientist. “Now biomedical engineers drive the research.” AI in liberal-arts and professional disciplines is like those early days of bio-medical engineering, he says. “This is why the gains in AI are still ahead of us.”Emory has been recruiting those “bilingual” talents steeped in AI and other fields for research into privacy and copyright law, personalized medicine, digital pathology, marketing and business, and other disciplines, he says. Just as engineers a few decades ago cross-trained in biomedical fields because they felt that’s where the future was (Emory’s Philip Santangelo, whom we met in the Robotics, Biotech, and Health section of this report, is a good example), a whole cadre of computer scientists wants to apply that expertise to other fields, Bellamkonda says. That has the additional advantage of bringing more women and underrepresented minority scholars into STEM and STEM-adjacent fields.The same interdisciplinary trends are reflected in student interests and curricula, he says. “In many ways our young people are ahead of us. They’ve been using these tools for a long time.”As human beings, we should be insulted by the notion that AI is a power so daunting that we have to hide it away from ourselves, says Bellamkonda. “There is a real difference we can make with these technologies.”In the medical domain, AI-enhanced digital pathology helps identify cancerous cells. AI-aided identification of cell-cycle proteins can help predict which therapy and timing might work best for a particular breast-cancer patient, and what kind of side effects the patient might experience.Of course AI introduces issues of privacy, security, bias, and malevolent use that researchers should focus on, he says, “but that doesn’t mean you don’t pursue it.”To pull back for a moment from the dystopian alarms and utopian cheers, it might help to remember what AI is. In its simplest form, an IBM primer instructs, artificial intelligence is a field that “combines computer science and robust data sets to enable problem-solving.” University researchers have been doing that for decades, and that work is accelerating rapidly. Let’s look at some examples.Medical Triage and ProtocolsA new protocol has reshaped national lung-transplant allocation through a system based on AI and machine learning. It was developed in part by Dimitris Bertsimas and Nikolaos Trichakis, faculty members at MIT’s Sloan School of Management, to make the distribution of donor lungs more equitable, reduce overall deaths, and increase access to transplants. Their recently published study was written with an MIT doctoral student and policy analysts from the United Network for Organ Sharing (UNOS), the nonprofit organization that oversees the transplant waiting list.More than 110,000 patients are on waiting lists for organs. More than 40,000 received transplants in 2021, about 2,000 of those lung transplants. Under the old system, candidates were placed in priority tiers, and some patients’ circumstances placed them at the boundary between tiers based on geography, age, gender, race, and blood group.The new model works on a point system based on medical urgency, nearness to a donor hospital, and other factors, resulting in fewer disparities related to age, gender, and blood type, and projected to result in fewer deaths of patients waiting for transplants and fewer deaths post transplant. AI and machine learning contributed to the analysis of trade-offs in patient outcomes. Following the new protocol’s initial success, Bertsimas says, more MIT Sloan students will work with UNOS on a similarly revamped protocol for kidney donations.A recent study by researchers at Yale and in Brazil, England, and Greece yielded an AI-powered triage platform to help contend with viral outbreaks. The team used machine learning and metabolomics — the study of small molecules relating to cell metabolism — to allocate resources in overloaded health-care systems.The team collected data from 111 Covid-19 patients admitted to Yale New Haven Hospital during a two-month period in 2020 and from 342 healthy people as a control group. Researchers identified a number of metabolites correlated with disease severity. They note especially that patients with an elevated count of eosinophils, a type of white blood cell, were found to have a worse prognosis, and patients who required positive airway pressure or intubation had decreased levels of plasma serotonin, a chemical that carries information between nerve cells.The triage platform, a user-friendly software, helps clinicians predict disease progression and severity, and estimate period and type of hospitalization, allowing for quick treatment to optimize outcomes. While tailored to Covid-19, the study’s authors say, it lays the groundwork for handling other kinds of viral outbreaks, too.A Combined Robotics-AI ChallengeWhen Navy ships come to port, maintenance and security teams demand inspection of their hulls. That’s labor intensive, requiring substantial diver hours, and it’s difficult and potentially dangerous, with bad light, wind, currents, and the possibility of mines.Sounds like a good job for a robot, and the Office of Naval Research gave Brendan Englot, as a young engineer, an opportunity to explore that. “It was the ultimate robotics challenge,” he says, and it inspired him to study robotic navigation more generally.Englot is an associate professor of mechanical engineering and director of the Stevens Institute for Artificial Intelligence at the Stevens Institute of Technology. He develops machine navigation systems for “stochastic” — random, or at least seemingly random — conditions, throwing in the added complication of no GPS. Boats or submersibles in stormy waters. Vehicles in chaotic traffic patterns or on remote, rugged terrain. The approach he’s taking is called “distributional remote learning,” which uses sensors and AI to teach a boat, car, or other machine to safely get around.His recent research involves simulations of self-navigating cars making their way through unpredictable heavy traffic and pedestrian crosswalks, or boats in waters with inconsistent currents and obstacles, as in whitewater rafting.Engineers and computer scientists teach robots through rules and reward-point systems. But how do you reward a robot? Lest the discussion drift into Philip K. Dick sci-fi reveries about robot self-awareness, Englot points out that a robot is what it does, and it does what it’s programmed to do. “At the end of the day,” he says, “it is remarkable how well this reinforcement-learning paradigm works.”Then again, these are simulations. Gradually, Englot and his colleagues hope to start testing these machine-learning systems on real vehicles.Imagine an autonomous rescue raft coming to your aid in a remote section of the Colorado River, or your car finding its careful way in a blizzard whiteout on a mountain highway where you’ve lost satellite connectivity.Human and AI CopilotsRecall in Section 3 that Howard University was mentioned as the first historically Black university likely to cross back into the Carnegie R1 research classification.One reason for that is RITA, or the Research Institute for Tactical Autonomy, for which Howard is the lead institution among a consortium of nine historically Black colleges and universities.The goal of RITA is to research and develop trustworthy AI capabilities for the U.S. Department of Defense so that fighter pilots can work with robotic copilots to defeat enemy combatants. The U.S. Air Force is committing an initial $90 million over five years to the university-affiliated research center, or UARC — the 15th such center funded by the Defense Department and the first under the Air Force.HBCUs train 30 percent of Black scientists but receive only .05 percent of the Defense Department’s almost $ 100-billion R&D budget. Legislators pushed for HBCUs to get more, and Defense limited competition for RITA to the 11 HBCUs that are vying to become R1 institutions. Strengthening the STEM pipeline to and from HBCUs is a key component of the planning.“I’m told this is the largest research award ever received by an HBCU,” Victoria Coleman, chief scientist for the Air Force, told Science. “The underresourcing of HBCUs didn’t get created overnight, and the problem isn’t going to get solved overnight. And in my book,” this research institute “is still not enough. But a vehicle like UARC, something that’s going to be there for the long haul, can be an important stepping stone.” Danda B. Rawat, an electrical and computer engineer and engineering associate dean for research at Howard, is RITA’s founding executive director. He tells The Chronicle that initially, the project will include faculty members, research scientists, and graduate and undergraduate students, as well as 14 supporting administrative staff members. Although the Air Force is funding the project, a priority will be creating training and communication technologies that can be coordinated among the domains of air, space, land, water, and cyber for all military branches, he says.RITA will explore how to foster trust between human and AI in training, reasoning, flight, and combat — with the human in command in some instances, and AI in command in others. Humans and AI have complementary skills and abilities, Rawat explains. A human pilot might be better able to analyze a situation, while AI might have better sensor-garnered visual and other information. Humans can read each other’s faces to detect problems, uncertainty, tension, worry. AI pilots might learn to do that too, he says, while also measuring their human copilot’s pulse, blood pressure, perspiration, and other stress indicators.Asked if there might be campus resistance to a high-profile military project, Rawat says he doesn’t think so because, unlike bomb-making or other weapons design, RITA will serve as a deterrent to war. And, he says, “if we don’t do it, someone else will.”Aside from the technical challenges of human-AI copiloting, there are issues of accountability and ethics. We know humans face consequences if they mess up. Would their AI copilots? In military and crossover civilian applications of such technology, how would insurers treat combined human-AI risks? What kinds of certifications and degrees might ensure quality training and operation? Those issues, too, will be studied in the RITA project.As far as fostering trust between human and machine, in military and civilian technologies to come, Rawat speculates that the key might be options for more or less automated involvement, the same way you can turn off your video in a FaceTime or Zoom call, or start and stop cruise control, GPS navigation, or auto-steering in your car.The Genesis of a More Human AITimothy Beal is a professor of religion at Case Western Reserve University, but his conversion story might not be the kind you’d expect.About seven years ago, he joined some other humanists in playing around with computer programming in Python. He was just curious, and rather than moaning about Big Data stealing our identities, privacy, attention spans, and intellects, he thought he should actually see what the whole thing was about.He liked “just to play with the tools and see what new types of questions emerge for us. And it really ended up being transformative for me, kind of a second life in my academic career.”He and some colleagues started h.lab — or “humanities lab” — to bring computational power to the humanities and humanistic perspectives to computation and to online, augmented, and virtual realities. The lab’s projects include developing a course in responsible AI and data citizenship that it planned to offer in 2024.Beal is also applying his new machine-learning interests to his background in the Hebrew Bible and its translation. AI usually involves filtering hidden large-language-model data sets into a single output. In contrast, Beal is experimenting with small language models that are not hidden, to explore the rich layers and alternatives in the meanings contained in biblical verse.In the beginning of Genesis, for instance, did God create the heavens and Earth out of nothing, or out of chaos? That may depend, Beal explains, on how you translate the biblical Hebrew word bereshith. Is it the noun beginning, like “In the beginning,” or the verb beginning, like “In beginning …”?If the latter, “this is a story that begins not at some absolute metaphysical beginning, with nothing, but in the midst of things, with God creating, or rather giving shape and form to things that are already there in some sort of primordial, formless form,” Beal writes in a chapter of The Routledge Handbook of Translation and Religion.Do you really want AI to decide for you your conception of creation or cosmology? Or do you want options and layers and a range of interpretations beckoning to you from a dynamic page?We think of AI as concerning large data sets filtered and tidied into a quick answer. Beal would like more humanists to explore AI’s potential to examine smaller but more-diverse data sets for analysis of the local and the particular.That Human TouchHow do you feel when you interact with AI?That’s what Gizem Yalcin Williams, an assistant professor of marketing at the University of Texas at Austin, and three co-authors explore in a 2022 article in the Journal of Marketing Research.Nuances are lost in summary, but overall they found that people feel better receiving favorable decisions from humans than they do when receiving them from AI. That’s because they find it easier to take credit for human-made favorable decisions, like when their membership applications or loan requests are approved. When they get bad news, however, like getting rejected, they feel equally crummy no matter whether human or AI informs them, and will blame whoever, or whatever, made the decision.That is among many AI-response questions emerging in our era, Yalcin Williams explains in an interview. For example, do you feel offended if Netflix’s algorithm suggests a movie or Spotify’s recommends a song that you find in bad taste? What about your choices and habits so far, you might ask, would make the algorithm think so little of you?How, too, are interactions with AI modifying our behavior? “We are entering this era when we don’t even have to produce full sentences,” Yalcin Williams says. Just grunt the name of a show into your smart remote and it’ll fetch it for you. Or just say “time,” and Alexa will tell you. No “please” or “thank you” needed.Alexa may not be offended, but what if you start talking that brusquely to your family, friends, and colleagues? “There’s a generation growing up and even more generations living with these technologies. Yet, we know so little about long-term consequences of interacting with AI,” Yalcin Williams says. Some companies even started proactively tackling such concerns. For example, Alexa has a function that rewards kids who say “please” and “thank you” with gracious replies.The long-term consequences of generative AI platforms like ChatGPT are also unknown. Will we become lazier and let them work for us, or will we use them to improve our skills and focus on tasks that are important to us? As it goes with any other technology, AI is a tool. How will we wield it? “You can use a hammer,” Yalcin Williams says, "to build a house or to demolish a house.”Quantum ComputingPeter Shor writes poetry inspired by the works of the Game of Thrones author George R.R. Martin, whom he superficially resembles — unruly white hair, mustache, beard, and a thoughtful mien.But Shor, an applied mathematician at the Massachusetts Institute of Technology and winner of a 1999 MacArthur genius grant, is best known for Shor’s algorithm, which he developed in 1994 to find prime factors of integers. Like one of Martin’s future-glimpsing “greenseers,” Shor was ahead of his time. The computer that could make optimal use of his algorithm doesn’t exist yet — and probably won’t for decades.That’s what happens when you contribute to a field in its toddler stage, and that’s where quantum computing is — more than a dream, with vast potential, but still trying to find its legs and prove itself.In 1900, the German theoretical physicist Max Planck hypothesized the “quantum” — the smallest discrete unit of a phenomenon. His hypothesis proved correct, and scientists since have identified numerous quanta, like photons in light and electrons in electricity. In the 21st century, scientists are taking those quanta and building increasingly sophisticated machines around them — computers but also sensors and information-transmitting networks.For quantum computing, qubits — short for quantum bits — are the basic unit of information, the way ones and zeros — digital signals that are either on or off — function as bits in classical computing. Qubits exist within electrons, photons, or ions (an atom with a positive or negative charge) when their intermediary states (“superposition”), together with interaction between multiple qubits, create mutually dependent states (“entanglement”).Using lasers, ultra-cold temperatures, and vacuum chambers, quantum-computer designers line up quanta and, with incredibly short light pulses (some a few femtoseconds, or one quadrillionth of a second), initiate a dance of the quanta. Because superposition and entanglement offer a much wider array of states than the binary zero and one of classical computers, and because quanta move so quickly, quantum computers can, in theory, solve much more complex problems rapidly.Popular culture has gotten way ahead of actual quantum advances. Johnny Depp’s brain was uploaded into a quantum computer in the 2014 film Transcendence, and in a 2023 episode of Black Mirror, a quantum computer creates a parallel-reality TV show with misbehaving doppelgängers of the actors Annie Murphy and Salma Hayek.Quantum computers do, in fact, exist — you can rent time on one for between $1,000 and $5,000 an hour — but, sadly, they can’t upload brains or create new Salma Hayeks. What they may someday offer is astonishingly fast calculations in pursuits like simulating complex molecules and their interactions, useful for drug discovery and industrial innovation; designing optimal traffic, supply-chain, and logistics systems; and encryption.Code Making and BreakingIt’s the encryption possibilities that have quantum scientists and quantum mathematicians like Shor so focused on prime factoring. But why? Did we really need a group at IBM in 2001 to use a quantum computer to calculate that 15 is divisible by 5 and 3?Wouldn’t a freelance second-grade been cheaper?We actually did need it — as proof of concept to show that the algorithm works. Because, as Shor explains, factoring a two-digit number is one thing. If you ask a classical computer to factor a 500-digit number, your great-great-great-great-great grandchildren will still be waiting for an answer. Quantum computers, when they advance, may someday yield that result within hours.Cybersecurity protects everything from online banking to gasoline pipelines to missile launches. So the right algorithm with the right computer, cracking 600-digit numbers in commonly used encryption systems, might someday win hot or cold wars. Conversely, lagging in that technology might lose those wars. No wonder the U.S. government is pouring money into quantum computing and related sciences.Meanwhile, it’s good that Shor’s algorithm was ahead of quantum technology, explains Shor’s MIT colleague William D. Oliver, director of the university’s Center for Quantum Engineering. That’s because Shor’s algorithm, and a recent competing one by Oded Regev, a professor of computer science at New York University, can be used now to improve current and recent encryptions that might be vulnerable when quantum technology advances.Cryptographers have to think many moves ahead like good chess players. Moreover, that chess game is taking on three-dimensional characteristics in a national-security push toward “post-quantum cryptography” that uses mathematical lattices, a type of abstract algebra. So cryptography going forward may be a mix of quantum and post-quantum cryptography.Tiny Particles, Huge PossibilitiesThe largest existing quantum computer, IBM’s Osprey, has 433 qubits, more than three times as many as its predecessor. But the quantum computers that could handle large integer-factoring calculations would need millions, maybe billions, of qubits.Moreover, environmental “noise,” physical interferences in these incredibly delicate processes, leads to small but unacceptable error rates that also limit quantum computing’s current utility. The combined scaling and error problems are why advanced quantum computing is still thought to be decades away.There are several predominant models, or “platforms,” for designing quantum computers. And scientists, for computing and other quantum applications like sensors and transmitters, are experimenting with a range of particle types, semiconductor materials, and laser pulses to hold ions still and generate ion interactions.Oliver’s research group at MIT focuses on solid-state quantum computing, in which qubits interact in a microchip supercooled to just above absolute zero (-459 degrees Fahrenheit). The cooling is necessary because heat increases noise, or error. But, Oliver says, the MIT Center for Quantum Engineering overall includes research on qubits from ions, neutral atoms, semiconductors, photons, and nitrogen-vacancy centers, which are photoluminescent point defects in diamonds.Christopher R. Monroe, a professor of electrical engineering at Duke University, formerly at the University of Michigan and then the University of Maryland, works with the trapped-ion platform in which atoms are held by laser cooling in a vacuum chamber the size of a deck of cards. “An atom is an atom,” he explains. “Standard. You can replicate it with absolute perfection.” Then the trick is to manipulate the line of atoms using lasers.That approach is being commercialized by a company called IonQ, for which Monroe is co-founder and chief scientist. In less than a decade, he says, it has grown to 300 employees in College Park, Md., and Seattle, with an office to open soon in Basel, Switzerland. IonQ has created eight or nine generations of commercial systems, some with up to about 30 suspended ions, he says. A couple of those systems are already in commercial service through the cloud, and the company is teaming up with Hyundai to simulate reactions with lithium compounds toward creating better lithium batteries.MIT’s Oliver says the best analogy for the state of today’s quantum computing is the Wright brothers’ 1903 flights at Kitty Hawk, N.C. No one looking at the wobbly craft reaching 852 feet at no more than 10 feet in altitude would have said, “Ooh, I want to fly to Hawaii on that,” but they might have foreseen distant commercial possibilities. Consider that 14 years later, by the time the United States entered World War I, thousands of planes had been built, and today, on peak air-traffic days, at any given hour there are 16,000 planes in the sky.While quantum computing generates understandable excitement, researchers say other aspects of quantum science might have significant impact long before quantum computers do.At the University of Michigan, Steven Cundiff, a professor of physics, and Mack Kira, a professor of electrical engineering and computer science, direct the new Quantum Research Institute, which is part of a $55-million universitywide endeavor designed to strengthen collaborations between faculty members, government, and industry. In 2022, the two also helped create the Midwestern Quantum Collaboratory, which has intensified collaboration between the University of Michigan, Michigan State University, and Purdue University.The institute and Collaboratory together, Cundiff and Kira say, involve about 100 researchers from the natural sciences, engineering, and economics.They say that very precise quantum sensing and energy-efficient quantum-networking technologies may develop more quickly than quantum computers. For instance, quantum sensors might detect light or gravitational waves that have evaded detection so far, or subtle air or ocean currents. Or by combining quantum interactions with atomic clocks, which keep time by the electromagnetic frequencies of cesium, rubidium, or other atoms, researchers might be able to devise supersensitive motion detection that would allow for navigation systems without GPS.Picture going through a long tunnel in your car, working in an underground mine, or spelunking through caves, and your electronic map still showing you exactly where you are.A Final WordWe’ve taken a hard look at the challenges and opportunities facing today’s research universities — tensions over funding for basic versus applied research, debate over collaboration with China, competition for elite classifications and associations, efforts to amplify the regional impact of university research, and scientific exploration in four particularly intensive areas. Watch, in coming years, for developments in all of those arenas.But obscured sometimes in the policy and strategic discussions are the everyday moments in which novice and seasoned researchers alike fan and direct the flames of their curiosity. That entails extensive knowledge and diverse skill sets, but also a patient, disciplined temperament and frame of mind.“I have not failed. I’ve successfully found 10,000 ways that will not work.”Thomas Edison said that — or something similar. And like so many researchers, Bryan C. Eigenbrodt, the associate professor of chemistry at Villanova University whom we met in Section Five’s exploration of self-cleaning fuel oxidation, appreciates the sentiment.Eigenbrodt loves working with undergraduates and igniting their research passions — wowing them, for instance, with a trip to perform an experiment at the particle-accelerating Advanced Photon Source at Argonne National Laboratory, near Chicago.But the students he enjoys working with the most aren’t necessarily those who elegantly navigate textbook lab exercises — because those are recipes for experiments others have already perfected. They build false confidence, he says. Real science isn’t like that.Students will need to create a different mind-set when conducting independent research. “The greatest thing I can teach my students is to fail,” he says — trying something new and pushing the boundaries of science. As with Edison, failure is the natural progress toward success as long as students take the opportunity to learn from their failures.For all that, he says, research “is addicting. The lows are really low, but even small successes can feel very high.”The experiments conducted in the Eigenbrodt Research Laboratory provide students with a “creation to application” experience. For example, when his students synthesize a fuel-cell catalyst, they incorporate it into working devices to generate voltage and current. And when it does, Eigenbrodt can see his students come alive, like Tom Hanks making fire in Castaway.“I can see it in their eyes,” Eigenbrodt says. “‘Look at what I have created.’”Return to Table of Contents

The Research-Driven University

Introduction

More News

From The Review

Upcoming Events