Skill Issue

For Meta’s artificial intelligence team, 2025 was a long, hot summer of hiring. In June, Mark Zuckerberg announced that the company had acquired a 49 percent stake in the startup Scale AI for $14 billion, bringing wunderkind and Scale AI founder Alexandr Wang into the Meta fold as chief AI officer. Soon after, Zuckerberg added industry veterans Nat Friedman and Daniel Gross and unveiled the new Meta Superintelligence Labs. In July, Meta raided competitors like Google and OpenAI, recruiting two dozen elite computer scientists and AI researchers to bolster the Superintelligence roster.

Media coverage highlighted the astronomical pay packages that Meta had offered the newcomers, reportedly worth as much $1 billion over a period of years. But scrutiny intensified after a Meta employee leaked an internal roster that revealed a striking detail: nearly all the new hires, and most of the Superintelligence team, are immigrants. Though Friedman is American-born, Gross was born in Israel. Longtime Meta computer scientist Yann LeCun, one of the world’s foundational AI thinkers, is French. Of the 36 employees added this year, 28 are immigrants. Today, Meta’s 44-person unit counts just 11 Americans and 33 foreign-born members, 21 of them Chinese.

Meta, an American company with seemingly bottomless coffers, has decided that its best chance to win the AI race is to recruit largely from abroad. For the U.S., this marks both a vulnerability and an opportunity. At the highest levels of science, technology, engineering, and mathematics (STEM), market choices indicate that American talent alone is not enough. Yet the country’s openness to gifted outsiders, its rich mix of commercial and academic institutions, and its cultivation of excellence make it the premier destination for the world’s top STEM workers. As the U.S. builds an industrial future with AI at its core, defining a national talent strategy is as crucial as it would be for any private company. But just as this reality has become most evident, the political tide has turned against hiring foreigners. While President Donald Trump has spoken favorably about skilled immigration, much of his MAGA coalition is set against it. Resolving this tension is now vital to charting the country’s course forward.

Silicon Valley earned its moniker during the period in U.S. history when the foreign-born population was at its nadir. From the 1860s through the 1930s, immigrants made up well over 10 percent of the national population. But by 1957, when Fairchild Semiconductor, the company that inspired the term “Silicon Valley,” was founded in California’s Santa Clara County, the foreign-born population of the U.S. was just 6 percent. When Gordon Moore and Bob Noyce broke away from Fairchild to start Intel a decade later, the foreign-born share had fallen even further, to below 5 percent. Though the Bay Area has always been a gateway for immigrants from across the Pacific, the then-peripheral Santa Clara County was more homogenous. In 1960, just 8 percent of its population was foreign-born, and only 3 percent was nonwhite.

A new era of migration to the U.S. began with the Hart-Cellar Act of 1965. The law has a complicated and contested legacy, but one clear outcome is that the U.S. now attracts more than one-third of the world’s skilled migrants. Santa Clara County has drawn droves of them, making it one of the most diverse places in the country by share of foreign-born residents. For more than half a century, its tech sector has absorbed skilled immigrants and reshaped the region’s culture.

Mostly from East and South Asia, Silicon Valley’s post-1965 immigrants and their children have integrated into the tech sector’s professional melting pot, helping set an achievement-oriented tone that distinguishes the region. Kim-Mai Cutler, a former journalist for TechCrunch and a partner with the venture capital firm Initialized, is a Silicon Valley native who has watched the region and its people transform. “I grew up in a primarily second-generation Indian, Taiwanese, and Chinese immigrant community . . . the rank-and-file engineering labor force that made big companies possible,” she says. “They were an important part of making Google, Facebook, NVIDIA, Intel, Oracle.”

The numbers bear out her experience. A 2016 paper in The Journal of Economic Perspectives notes that “56 percent of STEM workers and 70 percent of software engineers in Silicon Valley were foreign-born.” In the latest census, foreign-born Asian and Pacific Islander immigrants made up more than 20 percent of the Santa Clara County population, with native-born Asian Americans accounting for another 10 percent.

“These are cultures that really prize meritocratic competition, math, and science education,” Cutler says. Nowadays, the South Bay municipalities of Palo Alto, Cupertino, Sunnyvale, and Mountain View form a loop, with the American-born children of foreign tech workers matriculating at ever more elite public high schools, moving on to prestigious colleges, and then getting jobs at the tech giants or startups just north in San Francisco.

This influx of STEM talent conduces to the country’s overall economic benefit. Of America’s ten most valuable tech firms as of August 2025, six—NVIDIA, Microsoft, Google, Broadcom, AMD, and Tesla—have an immigrant founder, an immigrant CEO, or both. By venturing to the U.S., these individuals encountered not only other talented people but also institutions that enabled them to flourish.

And the benefits extend beyond the top of the corporate hierarchy. Throughout America’s STEM ecosystems, skilled immigrants have amplified returns for decades. According to a 2022 National Bureau of Economic Research (NBER) paper, “immigrants contribute directly to 10 percent of innovation, and their indirect contributions, through the enhancement of natives’ productivity, explain 26 percent of innovation. Together, immigrants account for 36 percent of total US innovation.” The same paper found that immigrants generated 23 percent of all patents and 25 percent of the aggregate economic value of patents over the 25 years ending in 2016.

The result of post-1965 high-skill immigration has been stronger economic output and higher incomes. A 2010 study found that the influx of skilled immigrants in the 1990s alone added between 1.4 percent and 2.4 percent to per capita GDP. The trend has continued into this century: between 2000 and 2015, a 2019 paper reports, foreign STEM workers boosted native-born workers’ incomes by over $100 billion cumulatively.

Foreign-born college graduates bring skills to the country that make firms more efficient and productive. While critics argue that STEM immigrants suppress pay by underbidding native-born workers, the 2019 paper found that between 2000 and 2015, the addition of foreign STEM workers actually raised the average income of native-born STEM workers by almost 5 percent, thanks to the complementary nature of their skills and the positive innovation spillovers that result.

Compared with its global rivals, America shows a unique ability to attract, retain, and maximize the talents of newcomers. Yet its immigration policies fail to make the most of this advantage. Every major pathway for permanent or temporary high-skill immigration is currently clogged, capping the gains that the nation can reap from foreign talent and trapping would-be contributors in bureaucratic quicksand.

The main route for skilled foreigners—defined in policy as those with at least a bachelor’s degree—to enter the U.S. is the H-1B visa, which permits temporary employment in professional fields. The visa ties recipients to a specific employer and grants an initial three-year stay, with the possibility of a three-year extension. Capped at 85,000 annually for for-profit employers—a number unchanged since 2005—the H-1B lottery has in recent years attracted 300,000 to 800,000 applications.

Instead of ranking H-1B applicants by merit, the government selects minimally qualified applicants as though drawing cards from a deck, with no regard for their skills, market value, or national priorities. “The H-1B program does not bring in the most skilled talent, it puts workers at risk of exploitation, and it subjects American workers to unnecessary competition,” writes Jeremy Neufeld, an immigration researcher at the Institute for Progress. “Many of these problems are not a function of high-skilled immigration generally; they are specifically caused by the incentives created by awarding visas at random.” Corroborating Neufeld’s argument on the competitive effects of the lottery, a 2017 NBER paper found that the H-1B program edges out some native-born computer scientists from the field and deters some American college students from majoring in it.

Like many regulations, the employer-sponsorship process distorts outcomes by favoring those adept at gaming the system and those with the deepest pockets. “Outsourcing” firms have become pervasive; they secure thousands of H-1Bs annually and then supply workers as contractors to other companies. And while the Metas and Googles of the Valley have teams dedicated to getting visas, startups do not.

Prescott Watson, cofounder of Aerolane, a firm developing autonomous cargo glider planes, described the challenge of the sponsorship system as a “non-negligible disadvantage” for small firms like his. “We don’t have the bandwidth to navigate the complex and onerous visa process with people,” he told me. “Unfortunately, because we are an early-stage startup, we almost have to exclusively rely on talent that is already fully legal and ready to work in the U.S.”

Though the H-1B is structured for six years—three plus three—thousands of H-1B workers remain in the U.S. indefinitely on extensions, due to processing backlogs in the logical next step: the employment-based (EB) green-card program, which provides permanent residency and frees workers from employer sponsorship.

Fully 140,000 EB green cards are granted annually, divided among five categories. For STEM immigration, the most relevant are EB-1 and EB-2, which allow self-sponsorship for immigrants who demonstrate extraordinary ability or serve the national interest. Both are capped at 40,000 annually. Yet, as with the H-1B lottery, the EB process does not weigh merit beyond basic eligibility. Instead, qualified applicants receive green cards strictly in the order that their applications are filed.

The program’s most damaging feature, however, is the country cap. No more than 7 percent of green cards can go to applicants from any single country, leaving candidates from populous nations like India and China stuck in a queue that may never clear, while those from smaller countries breeze right through.

The third of the major pathways for STEM workers is the O-1 visa, with about 20,000 granted annually. The O-1 is for extraordinary talents, but, like the H-1B, it is temporary. Recipients are granted three years in the country, and annual extensions are available thereafter. The O-1 has no formal numerical cap, but it is limited in practice by evidentiary standards ill-adapted to the tech world’s ever-evolving performance indicators.

Given decades of economic data on the value of skilled immigrants to America’s vitality and technological edge, the randomness of the H-1B process, the arbitrary country caps in the EB program, and the rigidity of the O-1 all work to the detriment of U.S. technological progress.

The contributions of STEM immigrants to the nation’s economy and vitality reach far beyond Silicon Valley. From university research institutes to midwestern software firms to secret Beltway defense labs, thousands of foreign-born scientists and engineers do essential work largely out of public view but to much consequence. According to the National Science Foundation, nearly one-fifth of all STEM workers nationwide are foreign-born.

I spoke at length with one such immigrant, Ketan Khare, about his tribulations with—and ultimate triumph over—the immigration system. After earning his bachelor’s degree at India’s Institute of Chemical Technology in Mumbai, Khare came to the U.S. in 2007 on an F-1 student visa to pursue a Ph.D. in chemical engineering at Texas Tech University. America, Khare says, was attractive to him then—and remains so today—as “a country that serves as the technological, cultural, economic, and geopolitical epicenter of the world.”

In 2013, doctorate in hand, Khare left Texas Tech for Rensselaer Polytechnic Institute in upstate New York through an F-1 add-on program called Optional Practical Training to continue his research. Two years later, he secured a postdoctoral fellowship at Georgetown University, along with the coveted H-1B visa for skilled workers. At Georgetown, he collaborated with the National Institute of Standards and Technology (NIST), building a successful career in molecular dynamic simulation. Asked to explain his subfield in layman’s terms, Khare says that it involves “using supercomputers to understand material properties from a chemical perspective.”

This is work that redounds directly to America’s national interest. At NIST, Khare collaborated with U.S. Army researchers on ballistics, studying chemical modifications to materials that could absorb energy from artillery blasts and keep equipment functional and soldiers safe. “My goal was to apply for a green card,” Khare told me. “I really enjoyed my research and wanted to contribute to what I was doing . . . but one practical part of it, other than liking what I did, was that if I had enough impact on my field—and I have—I would be able to sponsor myself in a category called EB-1.”

With the H-1B limited to six years, Khare faced a looming visa and career cliff just as the Covid-19 pandemic hit. When his visa expired in 2021, he was forced to return to India, joining the hundreds of thousands of would-be immigrants stuck in interminable backlogs and daunting odds. After filing his EB-1 petition for extraordinary immigrants, he spent nearly a year marooned in Mumbai, waiting for the U.S. consulate to review his credentials—a stretch he recalls as one of “lost productivity and a complete inability to plan.”

Still, the EB country cap gave Khare an advantage: though an Indian national, he was born in Kuwait. For U.S. immigration purposes, that detail proved decisive. Instead of falling into the oversubscribed Indian quota, he was processed under the far shorter Kuwaiti list. That biographical fluke meant that his consular appointment was scheduled for September 2023, while other Indians who had filed earlier remained in limbo.

Scientific bona fides in order, Khare says that the appointment proved a formality—an “extremely smooth, very courteous” one, at that. He returned to the U.S. the following month with an EB-1 visa and long-sought permanent residency. By 2025, he had secured a position at the University of Arkansas, where he now works in the chemical engineering department.

For decades, politicians have wrestled with the idea of comprehensive immigration reform, but incremental steps available today could deliver better outcomes for America’s economy and national interest. Neufeld, the Institute for Progress researcher, estimates that the U.S. could boost the H-1B program’s economic benefit by 88 percent over a decade without increasing the annual visa cap. His proposal: replace the lottery with a compensation-based system.

Such a system would send clearer price signals into immigration sponsorship and reduce the uncertainty that employers now face. Firms willing to pay the most for foreign talent would secure the visas. As Neufeld states, “Most employers would have a clear forecast whether a wage offer will be high enough to secure a visa or not, bringing the number of visas demanded and supplied into balance.”

As Neufeld outlined in a report earlier this year, a pure compensation-based ranking would significantly change the makeup of each H-1B cohort. Ph.D. holders—those most likely to drive high-end innovation and generate economic returns—would receive 67 percent more visas than under the current lottery. Neufeld estimates the net benefit to the U.S. economy of such a shift at more than $100 billion annually.

A plan announced by the Trump administration in September to add a $100,000 fee to H-1B applications would raise the stakes for employers and, in theory, shift the skill composition upward. But its sky-high cost would make the visas uneconomical. A pure pay-rank system is a cleaner solution to the oversubscription of the lottery. Moreover, because the plan does not spare the cap-exempt university H-1B visa, it would all but close off that nonprofit pathway, thereby excluding STEM researchers like Ketan Khare. The administration’s new lottery experience-weighting effort, also announced in September, misses the mark, too.

Within the existing EB green-card framework, the most useful reform would be to eliminate the country caps. Even if the overall cap remained, removing the arbitrary 7 percent limit would place EB-1 and EB-2 applicants on equal footing, regardless of their country of origin. Restricting natives of India and China—countries that each account for more than 10 percent of the world’s population—to just 7 percent of green cards leaves talent untapped.

While revisions to the O-1 process in 2025 that include more AI-relevant criteria are an improvement, the system is still too thorny. According to Wired, tech applicants are being “put through the wringer.” In a May report, the magazine cited Silicon Valley immigration lawyer Aizada Marat, who described a client repeatedly denied an O-1, despite serving as “a board member of a national bank in their home country” and having “raised venture capital funding from well-known angel investors and VCs to start a company in the U.S.” Proper vetting is essential, but clear criteria and objective evaluation against a rubric would be fairer to applicants and reduce administrative costs for both them and the government.

Foreign students represent another class of immigrants. While U.S. primary and secondary education lags some other wealthy nations, the country’s universities remain at the top of the pyramid. These institutions and their research labs are national assets that should be cultivated as entry points for future economic contributors. More than one-third of science and engineering Ph.D. students are foreigners on visas; three-quarters want to remain after graduation. The U.S. could give them that chance in greater numbers.

Yet the economic case for foreign STEM workers is not enough to win the political dispute at hand. Since its inception, the MAGA movement has been animated by cultural claims as much as economic ones. Trump himself has been characteristically fluid on the issue. Most of his commentary has been expansionist in tone, including a suggestion that foreign graduates from American colleges “get automatically, as part of your diploma, a green card,” an endorsement in August 2025 of continued Chinese student enrollment, and the initiation of a new “gold card” program for high-dollar investors. This attitude is also voiced by MAGA coalition allies like Ohio gubernatorial candidate Vivek Ramaswamy and Elon Musk. But with influential MAGA voices like Steve Bannon taking the position that high-skill immigration is as destructive to national greatness as low-skill, an administration pivot against it is easy to envision.

To hear the strongest case for reducing high-skill immigration, I turned to the Claremont Institute, speaking with Vice President Andrew Beck and Senior Fellow Jeremy Carl.

Beck invokes economist Vernon Smith to make his argument, suggesting that attempts to bolster the nation’s talent quotient risk tearing at the social, cultural, and institutional fabric that makes America—and the West more broadly—special. “If the American market has these bounds on it—the people we have available, the resources at our disposal, the time, the place, the circumstances we’re in,” he told me, “we need to adjust to those circumstances, not change the boundaries, or introduce a foreign factor to it. We have proven this, over and over in the West: we will adapt to our circumstances, and we will find a way.”

Carl sees things similarly. “I think the bottom line is that right now, immigration of all types is just way too high to keep social cohesion,” he says. “Diversity, ceteris paribus, is not a strength, regardless of what the Left says. Unity and a sense of belonging is.” Carl’s position draws validity from the historically high proportion that the foreign-born population has reached this decade: about 15 percent.

Beck and Carl rightly call out the problems with the H-1B program and the gaming that it invites. And neither would rule out all top-notch immigrants—yet, plausibly, given the foreign influx, they argue that the U.S. needs to consider seriously the risks of cultural balkanization. “We don’t want to micromanage, obviously, but institutions have a role in the design of markets and in shaping participants’ perceptions of them. [Again, this channels Vernon Smith.] And culture must be factored in. You cannot see people purely in a utilitarian manner; there are too many other realities at work,” Beck says.

Can the STEM talent-accumulation case be harmonized with culture-forward immigration restrictionism? Neufeld’s emphasis on altering visa criteria and prioritizing genuine market mechanisms would certainly help. By shifting the composition of immigrant cohorts as far up the skill ladder as possible through bidding and a more country-neutral approach, America can reap greater benefits from immigration, without net numerical additions. More rigorous skill-based evaluation will also help further differentiate STEM immigration from the mass migration and asylum-seeking fiascos that have engulfed countries across the West.

Those urging the Trump administration to reduce skilled immigration outright often point to the success of early tech companies like Fairchild, Intel, and Hewlett-Packard as proof that native-born Americans can go it alone. Those firms did advance the technological frontier with workforces made up almost entirely of Americans—but they operated in a world where the U.S. had no true competitors. Europe, Japan, the Soviet Union, and China were still rebuilding populations and capital stocks devastated by war.

Today, America has a population of nearly 350 million, while its chief rival, China, has more than 1.4 billion people and an economy that is gaining ground. Since the post-Mao era, China’s talent strategy has been to tap its vast rural population and steadily educate more STEM students than any other country. As recently as 1980, only two-thirds of Chinese were literate; today, students in major metros routinely outperform global averages. In Beijing, Shanghai, Jiangsu, and Zhejiang—together home to a population that would rank eighth worldwide if it were a country—students lead the world in science and math, according to the Programme for International Student Assessment. At the higher-education level, China now produces nearly 80,000 STEM Ph.D.s annually, compared with about 40,000 in the U.S.

While China faces a demographic crunch, its advantage in raw numbers will endure for decades. What China lacks is America’s ability to inject new blood. Though it is aggressively recruiting scientists and technicians from its vast diaspora, it is an undesirable destination for most other people. Attracting global talent delivers returns that the data mask. As Aerolane cofounder Watson puts it: “There’s a huge talent pool here in the United States, but if you look deeper, you’ll notice that a lot of the best people are first-generation immigrants who have become American, or children of those immigrants. That’s not to say that a lot of multigenerational Americans aren’t hardworking, talented people. It’s just a simple result of the huge level of interest from high-achieving people all over the world to come and make a life here in the United States.”

Wang, tapped this summer by Meta as its new chief AI officer, is one such second-generation success story, the son of Chinese immigrants who worked at Los Alamos National Lab in New Mexico. He embodies what Cutler observed growing up: “There’s a lot of first-generation Chinese immigrants who are really important, who come here and their kids grow up and they’re pretty American.”

While the clustering of immigrants lacking English-language skills in ethnic enclaves of some parts of the United States (and even more starkly in Great Britain) has been a social disaster, filtering for education and tying immigration to high-paying employment tends to avert such problems. The now-diverse Santa Clara County is not a modern Babel, made incomprehensible to Americans by the influx of immigrants in recent decades. It is a healthy American suburban environment with strong civic engagement and social capital. It is a place where the sons of Asian engineers take to the gridiron under Friday Night Lights alongside teammates whose families have been in California for generations. The melting pot—here, at least—remains a powerful force.

With due respect to cultural concerns and the importance of assimilation, to compete with its larger rival the U.S. needs to accept some of the trade-offs that come with immigration in order to derive the innovation benefits that will help sustain the country’s technological supremacy.

AI is the field where harvesting global talent is most critical to sustaining the American project. Meta, Google, Anthropic, OpenAI, and other U.S. firms are advancing efforts that could shape the international balance of power—initiatives placed under American jurisdiction and guided by American norms. When the alternative is to see some of the world’s best tech thinkers working for DeepSeek in Hangzhou or Baidu in Beijing, even the calculated risk of hiring certain Chinese nationals in Silicon Valley may be worth taking. While Chinese students and workers present a challenging case, with sufficient screening and minding, supported by law enforcement, they could offer a double dividend, adding to the U.S. talent base while subtracting from China’s—a strategy that helped the U.S. prevail over the Soviet Union not so long ago.

Two hundred and fifty years after independence, the United States is not just a paper construct but, to the Claremont Institute’s point, a peopled nation with its own culture. In a competitive era, however, resting on cultural laurels and relying solely on domestic abilities would result in the country’s being surpassed. America today is the “technological, cultural, economic, and geopolitical epicenter of the world,” as Khare, the chemical engineer from India, puts it. But that status cannot be taken for granted.

To accelerate progress, outpace China, and preserve the privileges that come with its singular position, the U.S. must keep its doors and labs open to the top technicians, researchers, and engineers of the future. To harmonize this pursuit with MAGA’s valid cultural concerns and maintain political viability, it must do so with a stricter eye on quality and an emphasis on the melting-pot ethic that served it so well in the past.

Top Photo: Tech leaders like Meta’s Mark Zuckerberg tap high-skill immigrants for their workforces. (Jeff Chiu/AP Photo)