How should H-1B visas select talent and protect workers?

I think it's a good idea to have this 100K fee. And I'll tell you the reason why is because right now there's something like five times as many H-1B applications as there are slots. So I think they grant about 85,000 H-1Bs a year and many more apply for it. And as a result, they have a lottery where they just kind of, I guess, they randomly choose who the winners are going to be. And if you look over the past decade, roughly half the H-1Bs go to these IT consulting firms. And the average salary is like $65,000 a year. So it kind of puts the lie to this idea that you hear that H-1Bs are for like high skilled engineers, AI researchers, things like that. That's not in practice what happens. In practice, what happens is you have this lottery and a huge chunk of them end up going to low end IT jobs. And I think by putting this $100,000 fee on it, you encourage the applications to go to the actual higher skilled, higher paid jobs, where there's actually a shortage of Americans. And you encourage US companies to try to fill those jobs with Americans first. And so I think, you know, putting aside some of the details, I think the big picture here is that they're using market forces to put some scarcity around the H1B application. And I think what that's going to do is encourage applicants to actually be these higher paid, higher skilled jobs that the program is supposed to be for, instead of these lower end IT job shops.

Seek to 3:41

And so when you put all of this together, I think the sort of broad takeaway is from where this started and what it was intended to do, we've deviated pretty wildly. And I think that this is a very important reset. Now, the last comment I want to make is about the people that say, hold on, we are going to cut our nose off to spite our face, and it's going to stop an inflow of incredible talent. And what I would just remind people is that it is really important to remember that when you are in the United States for a Masters or PhD, you already get an automatic visa. It's called OPT. So you have multiple years when you graduate from a useful degree program in the United States to find a job. I have several of these folks that work for me at 80, 90 These are incredible grads from Carnegie Mellon. They are off the charts smart. But because they did a Masters or a PhD, they come with a couple of years, and you can oftentimes extend that. And that will give us a very good amount of time to figure out how exceptional they are. And then quite honestly, I would gladly pay the 100,000 to get these guys on an H1B program. So I think if we're going to try to return this to what it was meant to be, which is to help American companies excel, get the best of the brightest, these changes I think are very good measures to course correct and get us towards that.

Seek to 10:40

Well, I think there should be two separate programs for what we could call highly skilled workers. What you were referring to after the end of World War II, there was a secret US operation called Operation Paperclip where we tried to recruit German scientists and engineers. Between 1945 and 1959, America recruited I think 1,600 of these scientists. So it was both, call it disabling to an American rival or adversary, but also expansive because that was when the nuclear industry was growing and much of nuclear science was being pioneered in the earlier days in Germany. And so the kind of American workforce expanded, but more importantly, a new industry was able to be enabled and unlocked and grown in the US. And then the German state was disabled by losing these scientists. One could make the case that a similar sort of scenario should exist today, that we should have a second operation paper clip. And perhaps it should be a continuing process rather than necessarily kind of this laissez-faire process that we have today, where we identify some of the top industries and the top scientists in the top domains and go after those scientists proactively with government action, government support in partnership with private industry. If you look at papers being published across mainstream scientific journals, the majority of papers today across nearly every scientific domain are being published out of China. And this ranges from physics to chemistry to material science to biotech. And there's a real case to be made that perhaps those scientists would be better off, and America would be better off if they were doing their research, pioneering here rather than there. Yeah. So I think that there's a very good strategic case to be made that perhaps like a more directed, high energy, high effort kind of operation paperclip be undertaken again around the world. The H-1B program, I do agree, has been heavily abused as a way of kind of compensation arbitrage. And if you find a highly qualified, excellent talent, as we all know, for a high skilled laborer in engineering or science today, that person, if you amortize the H-1B over seven years at 100k, that application fee, that's 15k a year, call it, that certainly seems worth it for the right sort of talent. And it forces the question about, can this person be found in the United States or not? The alternative would be to force a higher salary range, such that you as a company are now basically being forced to pay a higher salary, which means you have to justify that this person is worth it to bring them in from ex-US, and you can't find the talent locally.

Seek to 14:00

I think that we have to overhaul the H-1B program. Last night at dinner, actually, Howard Lutnick explained how some of these abuses happen. It's really unfair, actually, how it works. What he described is that when the application window opens, for what is a very small number of H-1Bs, a company that has, you know, call it 300,000 employees abroad, will apply on behalf of all 300,000, because they're all roughly the same kind of employee, whoever gets it gets to come over. Now, if you're filing 300,000 applications, obviously you have a disproportionately larger chance than Friedberg's company or my company or your company, Jason, who's filing one, obviously. And so, when those kinds of things happen and you can now use the data to understand it, you have to fix it. So, that's one very material and obvious change we need to make right away, which is we have to allow American companies to find these folks and have it be very precise. The second thing is that we're introducing a price that each of these companies can pay for, so that then you can signal clearly the disproportionate economic value that that person can create and the fact that after all the effort possible, you can't find that person here. And that's why you're willing to pay $100,000, which is a non-trivial amount of money. So I think that when both of those two things, the $100,000 thing is introduced and the visa application abuse is fixed, I think that we will go a long way to cleaning up the H1B thing and putting ourselves back in a much better place. But right now, there's just a lot of abuse. And so the program itself is not working the way it should have.

Seek to 37:39

First of all, there's definitely abuse. Second, it definitely needs to be corrected. The reality is that some of the H-1B visa holders are being paid below market wages. Some of them are not going to super talented individuals or in the jobs that actually require a lot of skill. I had a bipartisan bill actually that would have reformed it, requiring paying a prevailing wage, requiring making sure that the categories actually were skilled categories. I don't love the blanket $100,000 fee. I think that that's not the best way to reform it, but if, partly because it puts an unfair burden on startups, it actually may hurt with talent. But if you want it to say, look, there's going to be some prevailing wage standard and we need reforms, I'm open to that.

Seek to 1:40

First of all, there's definitely abuse. Second, it definitely needs to be corrected. The reality is that some of the H-1B visa holders are being paid below market wages. Some of them are not going to super talented individuals or in the jobs that actually require a lot of skill. I had a bipartisan bill actually that would have reformed it, requiring paying a prevailing wage, requiring making sure that the categories actually were skilled categories. I don't love the blanket $100,000 fee. I think that that's not the best way to reform it, but if, partly because it puts an unfair burden on startups, it actually may hurt with talent. But if you want it to say, look, there's going to be some prevailing wage standard and we need reforms, I'm open to that.

Seek to 1:40

Well, I think there should be two separate programs for what we could call highly skilled workers. What you were referring to after the end of World War II, there was a secret US operation called Operation Paperclip where we tried to recruit German scientists and engineers. Between 1945 and 1959, America recruited I think 1,600 of these scientists. So it was both, call it disabling to an American rival or adversary, but also expansive because that was when the nuclear industry was growing and much of nuclear science was being pioneered in the earlier days in Germany. And so the kind of American workforce expanded, but more importantly, a new industry was able to be enabled and unlocked and grown in the US. And then the German state was disabled by losing these scientists. One could make the case that a similar sort of scenario should exist today, that we should have a second operation paper clip. And perhaps it should be a continuing process rather than necessarily kind of this laissez-faire process that we have today, where we identify some of the top industries and the top scientists in the top domains and go after those scientists proactively with government action, government support in partnership with private industry. If you look at papers being published across mainstream scientific journals, the majority of papers today across nearly every scientific domain are being published out of China. And this ranges from physics to chemistry to material science to biotech. And there's a real case to be made that perhaps those scientists would be better off, and America would be better off if they were doing their research, pioneering here rather than there. Yeah. So I think that there's a very good strategic case to be made that perhaps like a more directed, high energy, high effort kind of operation paperclip be undertaken again around the world. The H-1B program, I do agree, has been heavily abused as a way of kind of compensation arbitrage. And if you find a highly qualified, excellent talent, as we all know, for a high skilled laborer in engineering or science today, that person, if you amortize the H-1B over seven years at 100k, that application fee, that's 15k a year, call it, that certainly seems worth it for the right sort of talent. And it forces the question about, can this person be found in the United States or not? The alternative would be to force a higher salary range, such that you as a company are now basically being forced to pay a higher salary, which means you have to justify that this person is worth it to bring them in from ex-US, and you can't find the talent locally.

Seek to 14:00

It's been very difficult to find some of the best young minds to work at our startups, to your point Ro, because it has been gamed and the people that have perfected the application process have won the H1Bs. And I think that that's where these systems go off the rails because it should be, as you said, the really talented young men and women that moved to the United States, supported by an American company trying to do something ambitious. It should not be because you know how to apply multiple times through multiple shell companies.

Seek to 3:10

Yeah, I agree. And as you know, some of the outsourcing consulting companies, I don't want to go through all the names, but you know what they are, TIDAC Consulting or Cognizant or others, they've gamed the system and they get a bulk of those H1Bs. And that needs to stop and there needs to be actual talent. But I want to make a point, I was just in China, we had gone to a bipartisan delegation there, one-third of the AI talent is in China, according to a lot of the reports. And so I want some of those folks to come to the United States, so we can stay ahead of AI. So there are legitimate uses of the H1B program, and what we need to do is fix the program, so you can have the legitimate talent still coming to Silicon Valley and around the nation.

Seek to 3:42

I think that we have to overhaul the H-1B program. Last night at dinner, actually, Howard Lutnick explained how some of these abuses happen. It's really unfair, actually, how it works. What he described is that when the application window opens, for what is a very small number of H-1Bs, a company that has, you know, call it 300,000 employees abroad, will apply on behalf of all 300,000, because they're all roughly the same kind of employee, whoever gets it gets to come over. Now, if you're filing 300,000 applications, obviously you have a disproportionately larger chance than Friedberg's company or my company or your company, Jason, who's filing one, obviously. And so, when those kinds of things happen and you can now use the data to understand it, you have to fix it. So, that's one very material and obvious change we need to make right away, which is we have to allow American companies to find these folks and have it be very precise. The second thing is that we're introducing a price that each of these companies can pay for, so that then you can signal clearly the disproportionate economic value that that person can create and the fact that after all the effort possible, you can't find that person here. And that's why you're willing to pay $100,000, which is a non-trivial amount of money. So I think that when both of those two things, the $100,000 thing is introduced and the visa application abuse is fixed, I think that we will go a long way to cleaning up the H1B thing and putting ourselves back in a much better place. But right now, there's just a lot of abuse. And so the program itself is not working the way it should have.

Seek to 37:39

Cognizant is not an American business. Infosys is not an American business. Tata, Wipro. It's not to say that that in and of itself is wrong, but you need to find the right visa class to do this under.

Seek to 10:26

Yeah, these are supposed to be for highly specialized workers. I can tell you, you know, when I was in IT in the early 90s, the abuse was happening all the time. And it was indentured servitude. It was disgraceful. The IT people would hire, typically Indians, and they would say stuff to the effect of, these guys are going to work for half as much and twice as long. And they can't say no. That's the best part of it. They can't say no when we put them on weekend coverage. They can't say no if we want to do a build out and they have to work 10 days in a row because we can kick them out of the country and they have 30 days to find a new job. And so it's a giant scam on the bottom half of these. I witnessed it firsthand. Every discussion I've ever had about H1Bs in relation to IT and consulting has always been about saving money. And the truth is, it's been abused. And I talked about this in 2015 on CNBC, Chamath, when Trump first started to talk about it. He's been on this for a while. And it's just great to see them. I had suggested 20K a year and that's kind of where they wound up. I additionally think they should do an auction for one-third of these. Let all these big tech companies that are truly trying to get in very unique PhDs from Oxford in AI, man, let them just put out how many they want to buy and at what price do a reverse auction and fill one-third of them with, I don't know, maybe OpenAI or XAI or Microsoft jumps the fence and pays $100K, $200K per person. What do you think Chamath is broadly speaking on this and the policy, the abuse, everything?

Seek to 5:19

Yeah. And the way to really do this, and this is Trump's superpowers, turning something that's a cost center into a profit center. I always give him credit when he does something brilliant. And the brilliant thing to do is to take the 100,000 and make it an auction. I would auction off half of these to the highest bidder. And then you would have Google, Facebook and Meta saying, instead of give me 100,000 of these at the rack rate, they would be saying, hey, I need 10 of these for sure. I'm going to bid a million. I need another 100 of these. I'm willing to bid 750 And then take that money and just allocate it to vocational training or retraining.

Seek to 40:35

So, but there's a process and that this process is going along, fixing the immigration system, right? Not letting people come in by lottery, selecting who they are, not bringing in people who would take a trainee job, right? People who graduate universities around the world, fixing the H1B. Yeah. I mean, look, I think H1B is fantastic for someone who makes $500,000, who's a cool engineer. But it's not okay if it's a $60,000 college graduate. Hire from the universities of America. Don't hire from the universities of the world. It just doesn't make sense.

Seek to 57:15

I think it's a good idea to have this 100K fee. And I'll tell you the reason why is because right now there's something like five times as many H-1B applications as there are slots. So I think they grant about 85,000 H-1Bs a year and many more apply for it. And as a result, they have a lottery where they just kind of, I guess, they randomly choose who the winners are going to be. And if you look over the past decade, roughly half the H-1Bs go to these IT consulting firms. And the average salary is like $65,000 a year. So it kind of puts the lie to this idea that you hear that H-1Bs are for like high skilled engineers, AI researchers, things like that. That's not in practice what happens. In practice, what happens is you have this lottery and a huge chunk of them end up going to low end IT jobs. And I think by putting this $100,000 fee on it, you encourage the applications to go to the actual higher skilled, higher paid jobs, where there's actually a shortage of Americans. And you encourage US companies to try to fill those jobs with Americans first. And so I think, you know, putting aside some of the details, I think the big picture here is that they're using market forces to put some scarcity around the H1B application. And I think what that's going to do is encourage applicants to actually be these higher paid, higher skilled jobs that the program is supposed to be for, instead of these lower end IT job shops.

Seek to 3:41

I think there needs to be an H1B program. I don't think the national waiver and the other programs are enough. A lot of times, look how Sundar Pichai came or Satya came, who are now leading Google and Microsoft. They came, they studied here, and then they got an H1B. But I would also make it that you don't stay on an H1B indefinitely, that we move towards a green card, which, by the way, President Trump said on your show, in one of your episodes. He said, I'm going to make sure that we actually move folks quickly to a green card. That to me seems like a win-win because you're then not exploiting an individual. You're going to pay them the market wage as they have flexibility of moving. And at the same time, they're going to stay in the United States instead of going back to China or to India. By the way, a lot of these companies, if you just limit H1Bs, they all have overseas, a lot of the big companies have overseas headquarters. So they're going to just offshore the jobs instead of bringing the jobs here.

Seek to 4:46

Yeah, these are supposed to be for highly specialized workers. I can tell you, you know, when I was in IT in the early 90s, the abuse was happening all the time. And it was indentured servitude. It was disgraceful. The IT people would hire, typically Indians, and they would say stuff to the effect of, these guys are going to work for half as much and twice as long. And they can't say no. That's the best part of it. They can't say no when we put them on weekend coverage. They can't say no if we want to do a build out and they have to work 10 days in a row because we can kick them out of the country and they have 30 days to find a new job. And so it's a giant scam on the bottom half of these. I witnessed it firsthand. Every discussion I've ever had about H1Bs in relation to IT and consulting has always been about saving money. And the truth is, it's been abused. And I talked about this in 2015 on CNBC, Chamath, when Trump first started to talk about it. He's been on this for a while. And it's just great to see them. I had suggested 20K a year and that's kind of where they wound up. I additionally think they should do an auction for one-third of these. Let all these big tech companies that are truly trying to get in very unique PhDs from Oxford in AI, man, let them just put out how many they want to buy and at what price do a reverse auction and fill one-third of them with, I don't know, maybe OpenAI or XAI or Microsoft jumps the fence and pays $100K, $200K per person. What do you think Chamath is broadly speaking on this and the policy, the abuse, everything?

Seek to 5:19

Yeah, I agree. And as you know, some of the outsourcing consulting companies, I don't want to go through all the names, but you know what they are, TIDAC Consulting or Cognizant or others, they've gamed the system and they get a bulk of those H1Bs. And that needs to stop and there needs to be actual talent. But I want to make a point, I was just in China, we had gone to a bipartisan delegation there, one-third of the AI talent is in China, according to a lot of the reports. And so I want some of those folks to come to the United States, so we can stay ahead of AI. So there are legitimate uses of the H1B program, and what we need to do is fix the program, so you can have the legitimate talent still coming to Silicon Valley and around the nation.

Seek to 3:42

I think there needs to be an H1B program. I don't think the national waiver and the other programs are enough. A lot of times, look how Sundar Pichai came or Satya came, who are now leading Google and Microsoft. They came, they studied here, and then they got an H1B. But I would also make it that you don't stay on an H1B indefinitely, that we move towards a green card, which, by the way, President Trump said on your show, in one of your episodes. He said, I'm going to make sure that we actually move folks quickly to a green card. That to me seems like a win-win because you're then not exploiting an individual. You're going to pay them the market wage as they have flexibility of moving. And at the same time, they're going to stay in the United States instead of going back to China or to India. By the way, a lot of these companies, if you just limit H1Bs, they all have overseas, a lot of the big companies have overseas headquarters. So they're going to just offshore the jobs instead of bringing the jobs here.

Seek to 4:46

I think the US has always indexed on human safety over speed to market and that has cost us in some ways. However, some other countries are indexed in the opposite direction which costs lives. There's a balance there. But there's research going on in China and other places, experiments and cloning and all things that, that as far as I know, aren't happening here. Yes, and I think the gutting of the CDC and the NIH and an anti-science vibe that has now pervades this country has driven a lot of mindshare elsewhere as funding is drying up for basic research.

Seek to 23:04

Yes, that as well. But the pushing out of H1B holders, there's so much happening now that it's causing... It's just easier to go elsewhere. That's not good for science.

Seek to 24:15

Well, I think there should be two separate programs for what we could call highly skilled workers. What you were referring to after the end of World War II, there was a secret US operation called Operation Paperclip where we tried to recruit German scientists and engineers. Between 1945 and 1959, America recruited I think 1,600 of these scientists. So it was both, call it disabling to an American rival or adversary, but also expansive because that was when the nuclear industry was growing and much of nuclear science was being pioneered in the earlier days in Germany. And so the kind of American workforce expanded, but more importantly, a new industry was able to be enabled and unlocked and grown in the US. And then the German state was disabled by losing these scientists. One could make the case that a similar sort of scenario should exist today, that we should have a second operation paper clip. And perhaps it should be a continuing process rather than necessarily kind of this laissez-faire process that we have today, where we identify some of the top industries and the top scientists in the top domains and go after those scientists proactively with government action, government support in partnership with private industry. If you look at papers being published across mainstream scientific journals, the majority of papers today across nearly every scientific domain are being published out of China. And this ranges from physics to chemistry to material science to biotech. And there's a real case to be made that perhaps those scientists would be better off, and America would be better off if they were doing their research, pioneering here rather than there. Yeah. So I think that there's a very good strategic case to be made that perhaps like a more directed, high energy, high effort kind of operation paperclip be undertaken again around the world. The H-1B program, I do agree, has been heavily abused as a way of kind of compensation arbitrage. And if you find a highly qualified, excellent talent, as we all know, for a high skilled laborer in engineering or science today, that person, if you amortize the H-1B over seven years at 100k, that application fee, that's 15k a year, call it, that certainly seems worth it for the right sort of talent. And it forces the question about, can this person be found in the United States or not? The alternative would be to force a higher salary range, such that you as a company are now basically being forced to pay a higher salary, which means you have to justify that this person is worth it to bring them in from ex-US, and you can't find the talent locally.

Seek to 14:00

Just on this operation paperclip point, it's interesting that China, the Chinese government took away the passports of the engineers at DeepSeek after the launch of that model, or at least it was publicly reported. I can't attest to this from firsthand knowledge, but there were definitely a lot of reports about this. And you can see why. I mean, if we could recruit or snap up a few hundred or at most a couple of thousand of these top AI engineers, that would be a game changer in the AI race. So China is hard to see those people as a strategic asset, and they're not going to let them immigrate, I don't think. But it would be something for us to think about. Certainly on the chip design front, there's probably just a few thousand people. That's all we're talking about. That would be a game changer on that side, although they're largely in Taiwan, not China. And again, I don't think the Taiwanese government is going to be too excited for us to snap them all up and move them to America. But in both these fields, there are a relatively small number of people, kind of like in the space race, who if they were all in America, would be a huge game changer.

Seek to 19:10

Here's the thing. We do have a really rich diversity of people from all around the world in higher ed institutions in the United States, getting masters and getting PhDs. We just need to be better organized about what to do with them. And we need to sort of reach out to those people, build relations with them, take advantage of OPT, and then we can always create a different class of visa for them. We have the ability to do these things called national interest waivers. So all of the infrastructure exists. And I think that if we can clean the decks on the H1B stuff, it will give people a lot more incentive to support the national interest waiver concept. I think the reason why people don't believe in this entire immigration conversation is on every part of the distribution of immigration, people see problems. They see an open border on the one side.

Seek to 20:21

And to just give people the history of this, this was something that was started after World War II to get really specialized people like Polish and German, like geniuses building out rockets. And I had a really interesting discussion, if you could pull this tweet up. There's another sinister wrinkle to this. I had this gentleman Colin on and he went to apply for a product manager. I had him on this week in Saratoa. It's my other podcast. And he applied and sent a resume with the reference number to the specific job, Friedberg, that he wanted at a company, New Relic. And he did this because in order to have an H1B visa, you have to put the job in a newspaper, right? So what these companies allegedly are doing is putting these jobs in these like obscure newspapers so that Americans don't see them. They're not putting them in places, you know, that you might see them. And there's a group of Americans who are going and finding these jobs and saying to Americans, go ahead and apply. Here's the shadow jobs, I think is what they call them. And so he put the reference number in there and they wouldn't even interview him. And I talked to him and he's kind of crestfallen. He's like, you know, I would like to apply for this job. But it's obvious that I can't get into it. This is like, I think just shows the entitlement of these tech companies. And I don't know New Relic's position on this. They can email us and I'll give it in the next episode. But they're basically listing fake jobs. And somebody, Abbey from PeopleOps over there, just kind of doesn't even let him interview for the job. The whole thing is just really dirty at the low end. And at the high end, it's under monetized. So, Friedberg, your thoughts on this? I know you have a lot of friends. You're an immigrant yourself. I'm not sure how you got here and what visa you came under. But I think it came when you were a kid, right? I'm not sure what your parents came under. But what are your thoughts on this and the impact it might have?

Seek to 12:06