Could you give us a quick primer on the mechanisms that have historically mattered most for enabling high-quality immigration to the US, specifically in STEM fields and technical training? What paradigms have been most valuable in the past? Then we can use that as a jumping-off point to dig into what has been changing over the last few weeks.

High-skilled immigration writ large is quite interesting because it's one part of the immigration debate that usually isn't politicized. It has broad support across the aisle, and it's pretty distinct from the kinds of immigration that are quite controversial, like illegal immigration. Historically, the US has not really changed its high-skilled immigration system for 30 years, so we've been relying on the basic program that has been in place for the last few decades.

Essentially, most of the people who come to the United States with high skills and high education, who have STEM PhDs, for example, we recruit mostly through international education. People start their journey through universities as students. We cast the widest net on that first part of the program, and then we have a big funnel where at each successive step, if people want to stay, we get more strict.

If people want to stay after they graduate, there's a program called OPT, Optional Practical Training, which allows them to work while they try to get another kind of visa. After that, they can try to get an H-1B, which is the largest high-skilled work visa. There are a couple of other programs as well. Most of the people on these temporary work programs started in school already. They're mostly used for people who are already here, not for getting new people from abroad. A small percentage of those people we let stay permanently on employment-based green cards, which is permanent residency in America. For those, 90% of the people are, again, already here. So international education is really the front end. That's our recruitment pathway. All of these other visas are best thought of as retention visas.

If you drill into STEM fields more broadly, you've talked about how something like 70% (60%) of top AI startups have been founded by immigrants who came on student visas. That's exactly the front end you're talking about. What are the implications of the current limitations being put on student visas, squeezing that top of the funnel, for technical fields that have relied on a steady flow of people coming in on student visas?

The key thing to keep in mind is that as roles require more experience and more technical expertise, we tend as a country to be more and more reliant on foreign-born talent. That's not too surprising, because we're not the biggest country in the world. China has four times our population, and for some years now they've already been graduating more STEM graduates each year. So we can do more to try to keep up.

But the big advantage we have is that we don't just have to rely on our smaller population, because many talented people from all around the world want to come to the United States, or at least traditionally have wanted to. We can tap into that and use it to our advantage. When we've historically done that, it's given us some of our biggest advantages. We got to the moon because we proactively recruited some of the top German scientists. More recently, if you look at top AI companies, for example, 60% of them were founded by immigrants, most of whom, as you mentioned, started as international students. By and large, our innovation system is disproportionately reliant on immigrants, because we can select the best and brightest from everywhere, and that gives us a huge edge.

That system is reliant on international education at the front end. We can talk about whether there's more we could do to diversify that, but it's a simple matter that today, our high-skilled recruitment program is essentially grad school in this country. If we're doing things that choke off the supply at that front end, we're going to get way fewer people down the road. That's going to have huge knock-on effects for innovation, for the development of new technologies, and for our competitive edge against countries like China.

One consistent thread in many of the current administration's policies over the last few months is an increased focus on and reliance on what's already in the US, whether that's manufacturing, existing supply chains, or existing trade. It's becoming very insular. One pushback from critics who want to limit this immigration is that there's intense competition for high-paying, high-skilled jobs in STEM fields. You've written about the fallacies in that argument: what is the true supply of highly qualified people, and are there shortages that undermine the narrative that we don't want to let people in from other countries because they compete with Americans who also want those jobs? How do you think about that push and pull?

I think there are two effects going on here. One is that the US doesn't have a monopoly on the world's talent. The superstars who drive new innovations, new technologies, and new business formation tend to be the top 0.1% of people internationally. We don't have a monopoly on those people, so the question is whether they're going to work here or abroad.

The second point I'd make is that, especially in scientific and research fields, there are huge complementarities. One of the ways immigrants can lead to US innovation is by inventing a new product themselves, but it turns out that the more common way is by combining ideas with people who are already here, and that combination ends up being more than the sum of its parts. So not only are they bringing ideas directly, they're making their native counterparts and collaborators more productive. It's not just that people come here and file patents on their own. The people they work with end up filing more patents as well.

So rather than thinking about this as a shortage, with some fixed number of jobs where we're just trying to fill slots, in which case I wouldn't think immigration is a huge part of the story, think about building an innovative ecosystem where people can invent new technologies and found successful new companies. There, our huge advantage is our ability to tap into people from around the world.

You touched earlier on the space program, where much of the success came from immigration practices that brought over German scientists. That's been a huge boon at different points in our history, and you've written about pulling off something akin to a modern Project Paperclip. Doing that today involves a lot of interconnected complexity. The pushback on immigration isn't just blanket limits on student visas for the heck of it. It's an attempt to break certain connections, for example, cutting off students with ties to the CCP as conflict increases. We're worried about foreign influence in our universities and about students with those kinds of connections, and there's a lot of gray area. How do you think about pulling off something like Project Paperclip today with those complex realities baked into the problem?

Historically, there's been a long tradition in American immigration policy, going back to before World War II, most famously with Operation Paperclip, where the US government proactively tried to recruit the top minds, the top scientists working on the kinds of projects critical for US security. I think the outcomes of that kind of program speak for themselves. That's how we were able to win the space race. I think that would definitely be worth doing, and it would be quite critical for technologies like AI, where getting just a couple of the absolute top minds can make or break our leadership in a new technology.

But I think that's quite different from what's going on with student visas, because it's very hard to tell who is going to be the top scientist in the world in the future, the superstar leading brilliant new discoveries. Even if we could successfully recruit people later in their careers whom we know are top scientists, programs like international education, where we cast a wider net at first and then let top people stay, give us a tremendous advantage. We get people earlier in their careers who are going to make their major contributions here, rather than trying to get them after they've already done something that identifies them as top talent.

I think the sense of scale here is important. A program like Project Paperclip 2.0, which is something I've called for, might only be 100 visas a year. International education right now is hundreds of thousands of Chinese students alone, and even more when you consider other countries. Significantly reducing enrollment there could pose significant challenges for the US science and technology ecosystem down the line that can't really be compensated for by just a few people.

One frustration is that people got excited about the prospect of stapling a green card to the diploma of every graduate of a top university. You've talked about how something like 10% to 20% of foreign graduates from top schools actually end up staying in the US, largely because of significant obstacles to doing so. What we wanted was a deeper unlock, with more people coming from abroad into our educational system and more of that talent kept here. What we're seeing instead is the opposite, shutting off that funnel. What are the driving forces behind that, and what are the biggest concerns? How would you respond to them in making the case that we should make it significantly easier for people in technical fields to stay after they graduate?

The fact that we're losing people is something the CCP has celebrated. One thing that's been funny to watch, funny and very sad, actually, is that a few years ago the CCP was expressing a lot of concern that the US would fix its high-skilled immigration system and improve its ability to retain people. That would pose a huge threat to them, because since the last party Congress, they've made a concerted effort to become a competitor to the US in talent recruitment by 2035. They know the way they might succeed in that race is by the US being complacent about its immigration system, which is exactly what we're doing.

Last year, when the president was on the campaign trail, he talked about stapling a green card to a diploma. Now it seems like some of his appointees want to expel those people instead. As a policy matter, one concern with stapling a green card to the diploma is that it can create a problem with visa mills, where universities don't necessarily have the right incentives to maintain their standards. Some other countries that have implemented a version of the staple-permanent-residency-to-a-diploma idea have found that their standards are declining. The universities end up seeing international students more as a cash cow. The US is lucky that we've avoided most of that, and part of it is because of the policy environment. So I think it's important that we don't create policies that end up reducing standards.

But at the end of the day, we know that most of the talent coming out of our top STEM PhD programs is foreign-born today. If there are no programs that can keep them here, or even worse, if there are programs that deter them from coming in the first place, those people don't disappear off the face of the planet. They go work somewhere else, and in many of those cases, maybe they're working against the United States' interests.

I think that's a perfect way to sum it up. Enabling that system is the way we bring that advantage here, and we're letting other fears and concerns get in the way and put a stranglehold on it. I really appreciate you hopping on to unpack this with us and taking the time to talk it through.

Yeah, thanks so much.