Hacker News Reader

Comment by ravila4 | Hacker News Reader

I feel like this thread has a lot of bias coming from bad experiences in the computer science field. While I do understand the bad practices of many tech companies, I think the picture is very different when you look at careers in other STEM fields, such as biology, chemistry, and physics. H1B visas are very important for foreign students in these fields particularly those hoping to pursue careers in academia. H1B visas are in fact a beacon of hope that brings many talented students to America, with the promise of finding a job here. As a matter of fact, many of these students go on to work on projects that are important for human health, e.g. COVID-19 research. It is convenient for the country to attract and retain their talent, of which there is currently a shortage. Not every H1B holder is an underpaid scrum master. Some are legitimately scientists doing on a postdoc in academia, and working to improve lives. For those of you rejoicing, I just want to ask you to do more research in how this rash policy change may impact the lives of real people.

View on HN

bias coming from bad experiences in the computer science field

https://www.myvisajobs.com/Reports/2020-H1B-Visa-Category.as...

It isn't bias, just look at the H1B distribution, it heavily biased to the CS/IT space. Like for every possible biologist you are imagining, there probably 100 IT/CS guys coming in. I'd love to have the biologists and chemists come in, but the big IT consultancies are massively abusing the system, and that needs to get fixed first.

Look at the top company sponsors, not STEM fields like biology, chemistry, and physics. It is entirely software.

Why? Are there more jobs for Biologists than Computer Scientists/Developers/Programmers?

It looks like this is a response based on demand. Why in the world would you want more biologists when there isn't market demand for them?

This is mostly wishful romanticization of particular "underdog" areas of studies. Makes no sense to me. Even more preposterous is the parent's profound claim:

> I feel like this thread has a lot of bias coming from bad experiences in the computer science field

I don't understand why we are talking about bias? It's not like the market goes "Gee ... we are a software company, but only if we would abandon our biases and hire an archeologist from Columbia and sponsor an H1b visa for them. Not sure what would do with them, but we must hire them though. Hey Rob! You got any ideas man?"

Market forces do a poor job driving science. Its too long term, too risky, and you lose to much to the tragedy of the commons. All but the most trivial and mechanical research mostly happens due to government funding, or as a prestige byproduct of higher education (ie, not directly driven by their profit objectives). So in this case, maybe there is demand in a social sense, not a market sense.

Alternatively, large computer science companies may have just greased enough pockets to create regulatory capture of the process, creating the inequality in a non market driven way.

> Market forces do a poor job driving science. Its too long term, too risky, and you lose to much to the tragedy of the commons

Then how did we end up with things like nuclear fusion startups with hundreds of millions of dollars in funding? Market forces do a superb job of driving scientific innovation. We live in the most scientifically innovative period of our history.

I find that what people who say this sort of thing actually mean is “the things I perceive to be valuable are insufficiently resourced (according to me)”, rather than “things that are valuable fail to find resourcing”.

> Then how did we end up with things like nuclear fusion startups with hundreds of millions of dollars in funding?

By millions upon millions of dollars and man hours spent by governments and universities doing basic research. Markets are very poor at laying the groundwork required for scientific breakthrough, but quite good at leveraging that space to their own advantage, once it exists. This is, in fact, why the US government used to spend so much on basic research, so that its Capitalist economy could advance innovation more quickly.

> We live in the most scientifically innovative period of our history.

You’d have to state your measure for this. Scientific breakthroughs are not advancing at anywhere near the speed of the early 20th century. This may be because we’ve picked off a lot of the easy targets in the fundamental aspects that describe our reality, but would be contrary to this sentiment regardless.

I don’t really get why you think this isn’t a market solution, and I don’t really see where you think the problem is at all. Research institutions compete for fee paying students, endowments, and government funding. When they come up with marketable ideas, those ideas tend to make it to the market.

The fact that private capital can even find its way to funding ideas that have not and may never be proven to work completely disproves your point. Long term, high risk business propositions don’t seem unable to find investors. There is no tragedy of the commons here. If the market values something, there’ll be somebody willing to produce it. The only real room for disagreement is when the market doesn’t value something as much as you think it should.

> You’d have to state your measure for this. Scientific breakthroughs are not advancing at anywhere near the speed of the early 20th century

There’s more to science than quantum mechanics and a good approximation of gravity. It’s generally seen as rather obvious that our rate of innovation is increasing exponentially. I’m typing this comment on a pocket sized computer that’s orders of magnitude more powerful than the most powerful supercomputer in the world was on the day I was born.

https://www.nature.com/news/polopoly_fs/1.19431!/menu/main/t...

But what we're slowly learning is that a huge percentage, perhaps most of that "millions upon millions of dollars and man hours" ends up producing terrible papers that don't replicate (in science) or don't even say anything coherent at all (e.g. basically all of gender/critical race theory).

COVID-19 has been an excellent demonstrator of this: at a time when it's become clear that epidemiologists and virologists don't actually seem to know much about viruses, and know even less about stopping them, we have widespread riots and protests against 'systemic racism' that can't be found in any data sources. All this badness is coming from universities which look bloated, over-funded and nowhere near scrutinised enough.

Then finally you look at the tech sector, you look at commercial organisations at the cutting edge. The sad reality is they don't use much university output. They don't. Look at CS, most of the breakthrough papers are coming from corporate labs. The last 10 years saw multiple new programming languages take off and gain critical mass, none of them bear any resemblance to Haskell or other research languages popular in academia. A lot of the improvement in tech, whether it's electric cars, self driving cars etc, it's been coming from corporate labs.

China has four times our population, and we are falling behind on many fields. We need all the manpower we can get. On another note, I'd like to raise the point that bringing talent to the country also creates demand. Many of the next generation of startups will be biology-based.

THe bias I speak of is in regard to the attitude towards H1B. I have my own different bias, for in my field most people are in favor of it.

It's not necessarily a demand issue. It may be that companies are abusing the system to gain more such workers. I've certainly been part of hiring searches where qualified US candidates didn't get the job. It often came down to money. I'm in a higher cost of living area, so salaries tend to be higher than average. But the company was offering lower than the national average, much lower than the local average. The candidates from the US were often offered the job, then turned it down because the pay was so low.

"I can't afford or don't want to pay market rate salaries" isn't supposed to be a valid reason to grant an H1B visa.

I think there is something weird going on. I am heavily involved in my company's staffing process, and the vast majority of the referrals we get from our recruiting sources are from overseas (and often cheat and have cruddy IT skills).

Absolutely. But perhaps it is precisely for this reason that we need a comprehensive and thoughtful reform that takes into account the value added. I am worried that blanket-raising the requirements solves one problem and creates another.

I think policy should be evaluated by whether it is better on net, not by whether it has any downside whatsoever. The latter is a sure way to be stuck in status quo.

The US needs way more software engineers than is has, and sky-high salaries are the evidence of this. I think we should make it easier for people to come and work in the US.

The existing H1B system works pretty well for employers and employees. When I hear people describe H1B immigrants as somehow taken advantage of, their experience seems far different than mine (I have never seen people mistreated by the system).

The idea that this system makes it harder for native software engineers seems crazy to me. I know many software engineers and both the great ones and the terrible ones have no problem finding work.

Does this mean the US needs way more CEOs and Board Members than it has, and the sky-high compensation packages are evidence of this as well?

No, programmers are underpaid, deliberately. We just had a story all about it. https://news.ycombinator.com/item?id=24689582

It has been suggested by Eric Weinstein that the H1B visa was deliberately designed to cap wages of PhDs... this is a side-effect.

Sky-high executive compensation is absolutely a sign that the US needs way more competent executives.

Restricting the supply of software engineers to the US will help software engineers at the expense of everyone else.

We just need to give the money to the people who actually create value, and the imbalance will sort itself out. Fair market capitalism is way better than the "Free" market.

The window to get kicked out of the country when layoffs happen is pretty narrow, even for people who have lived here for years. It presents a pretty heavy threat if abused because it might be hard to seal the deal on a new job (especially if you're even a little bit selective) before you suddenly become illegal.

From everything I've experienced in the tech industry the thing you've described is just hypothetical. Everyone I know on H1-B wages freely moves between jobs with no problems.

From the outside, it came seem like there are no problems. The process can be nerve-wracking. The interviews need to happen clandestinely because if your current employer finds out and fires you, you have to leave the country in 60 days. After weeks/months of interviewing, your new employer applies for a visa transfer. This process could take 3-4 weeks for application (again nerve-wracking if you are desperate to move), another 15 business days for the application to get adjudicated if you apply through premium processing. Under this administration, denials and Request For Evidence are being issued willy-nilly. So this 15 business days can end up being 30 or more. Again, the mental anguish can be hard if you really want to quit your current employer. Then 2 weeks after resigning. By the time you join your new employer, it could be 2-3 months. By law, one cannot take a break between jobs so you'll have to negotiate with your new employer if you need a break in between. H1B for people from India and China is indentured labor. Even as someone whose compensation is in top 0.5% in the US, it's extremely restrictive. I can't even imagine what the average H1B employee goes through.

In my experience, H1B system works pretty differently for employers and employees depending on whether you're talking about an actual tech company vs. the IT department of an industrial age company.

I fail to see the "problem" here. Most academia STEM jobs are incredibly competitive, postdocs less so than professorships but for fundamental science it is nearly indisputable that academia is flush with labour. If universities deem it too hard to raise postdoc wages across the board (or for specific foreign students) due to a lack of grant funding then they will still have ample choice of American students who go onto do the same research.

Yes, it is unfortunate for students looking to move to the US that the path through the standard academic route is no longer easily viable but that does not mean important research does not get done. There are also plenty of other countries that are in fact seeking such talent that might actually look like better options for the now disenfranchised (for lack of a better word) foreign students.

It's very simple. If it's harder for foreign people to become postdocs/profs in the US, the quality of US research will fall. Quickly.

> If it's harder for foreign people to become postdocs/profs in the US, the quality of US research will fall. Quickly.

I sincerely believe exactly the opposite. The best and brightest in the USA avoid research academia because the quality of life is terrible and the pay is below average. Like, "all my students make more in their first year than I make with a PhD and 20 years of successful research" is a common lament among CS profs everywhere outside of the top tier of R1.

Meanwhile, the "graduate school as a ticket into the country" incentive structure creates a lot of pathological behavior among the foot soldiers of academia.

Exactly this. I would have loved to go for a physics PhD but everything I read about academia in the US strongly discouraged me from pursuing it. I went into tech instead.

The academic system in the US is incredibly sick. The best and brightest students are certainly not lining up to squander their youth being treated like slaves when they can make top dollar in another field.

Citation heavily needed here. I find it hard to believe on face value that foreign researchers are necessarily better than US researchers; especially given the number of both foreign and domestic university graduates from U.S universities is much less than the available postdoc and professorship appointments in STEM.

A lot of the best academics are foreign born. You could look at Nobel prize winners for example.

But you have to also remember that the current group of academics came out of PhD programs in the 1970s. That was a time of expansion in academia. These folks are just about to retire. On the other side of this, I think the universities walk a fine line with respect to graduate student admissions between foreigners and US citizens.

The fact that we bring in foreign graduate students from foreign undergraduate programs either means that our current system is not producing applicants of sufficient caliber who then apply or it is not economically viable. In that sense there is evidence that graduate student research stipends are depressed. Is this due to foreign competition?

There is some anecdotal evidence to this in that the majority of CS undergrads get a job ratherhe than pursue a PhD etc... b/c of the salary differences. Foreigners have to get a masters at a minimum to achieve better odds in the H1B lottery. And after they graduate from a US masters program, they can work at a US company for a couple years anyway.

We're bringing in foreign students to graduate programs because colleges want to grow revenue and demand to go to grad school at all but the top 10% of schools isn't high enough to easily fill every seat. Outside of top tier programs, colleges are fighting for every student they can get, and international students are a good supply.

Ok but then why are graduate student "salaries" (e.g. stipends) essentially universally depressed? It seems like research programs should pay more to attract more students.

Because the supply of qualified applicants for ANY job is unimaginably larger than anyone imagines. Even if you're hiring assistants for the "best researchers in the world" -- which, let's be honest, are all doing the "10 years and retire" game at FAIR/Google/etc. anyways these days -- there are just unimaginably many qualified candidates.

Even killing the H1B program entirely wouldn't directly effect the ability of universities to recruit research/teaching assistants.

Consider if any US trucking line could hire literally anyone in the world. Wages would be below subsistence level. Which, well, lots of the Research Assistants working as Software Engineers for National Science Foundation portfolio companies are barely making enough to support themselves.

This doesn’t jive with the post I was responding to.

Because these wages are near universally low, so no one needs to compete on that. Also for the same reason they're recruiting international students in the first place: Money.

(Note: this doesn't mean the schools aren't fulfilling their educational missions, it just means that revenue is both necessary in itself and as such a core metric that is pursued.)

At least in my field, foreign students don't have H1B. typically F1. H1B is for postdocs and above. I think the whole discussion about foreign students is beside the point here.

Science is by far one of the most globalized fields. The international collaborations that it thrives on are largely formed via exchange of students. We cannot afford to be ethnocentric in this day and age.

As an example, America would never have been first on the moon if it were not for a certain German engineer Wernher von Braun.

There was no argument made for ethnocentricity, nor for stopping a global exchange of ideas. This response seems very non sequitur.

> The international collaborations that it thrives on are largely formed via exchange of students.

1) Students are on F visas, so they are unaffected by H1B changes.

2) There is no magical reason that students educated in another country in this day and age have a markedly different tradition of thinking about fundamental science. The internet has largely erased the differences in access to the latest research in top institutions around the world so the only thing that remains needed for high quality research is a pool of bright people. Perhaps there are some educational tradition differences between Europe and the US but US postdocs are largely educated in the US itself so those differences aren't really affecting current US research anyway.

3) Given the current labor and supply demand in academia, this pool of bright people can be satisfied by either US researchers or foreign researchers, and since there is ample choice between the two, the research of great import still gets done by just as many smart people as before, so once again, I fail to see the "problem" for great research being done.

There are, absolutely, tremendous differences in the way an American student/postdoc or an German student/postdoc attacks problems and thinks about science, speaking for the two countries I worked in.

I kind of agree, but von Braun didn't exactly come to us through "traditional" channels. The Russians were a little more forceful about it, but neither them nor the US gave the individuals a significant choice in the matter as we embarked on that scientist gold rush.

Quality is distributed somewhat evenly. The best and the brightest are all over the world. Not every foreigner is better then every local, but for any given quality cutoff, a majority of talent will come from outside the nation.

If your goal is sufficiency, then the local population will suffice. If your goal is to capture as much of the truly top talent, you have to cast the net wider.

I think we're making mutually reconcilable points here.

> Quality is distributed somewhat evenly.

Starting from that assumption (which I can agree with), here is a simplified (perhaps numerically inaccurate) argument. Assuming the top talent constitutes 1000 best researchers who have just graduated from their PhDs, and all relatively equally qualified, 500 US, 500 foreign.

There are 400 postdoc appointments. Currently if 200 postdocs are US and 200 postdocs are foreign and these H1B rules then inevitably disallow the next generation to have 200 foreign postdocs we still have 200 US postdoc candidates who are just as good and are still top talent.

Would this not still retain equal quality?

You put in the assumption that they are all relatively equally qualified. I don't think the 1000 best researches are equally qualified. The x best are better than the 1000-x below them. Additionally, if the exchange of postdocs is lower (many return to their country of origin), the research quality not only in the US, but world-wide will go down.

It’s just a numbers game. America has less than 5% of the worlds population. Why would you think we would have more than 5% of the worlds best researchers

It's a "problem" if you hear this and think, "Wow I want more of those people".

It's not a "problem" if you hear this and think, "Yeah, it should be hard for them".

Ultimately, people explicitly disagree with your initial position and framing.

Did you mean to respond to GP?

If not, that is a rather disingenuous reading of what I wrote. No one is saying it "should" be hard for foreign students to get a job in academia, my point was academia has ample domestic labour, so even if H1B salaries are raised, the research attempted is unaffected.

Wanting more researchers is well and good, there just aren't enough jobs and funding for them in the current STEM academic market (which is completely separate from any H1B related issue), hence the rather large number of people who move to industry.

> H1B visas are very important for foreign students in these fields particularly those hoping to pursue careers in academia.

Academia is a great example of how opening up a domestic job market to The Entire World suppresses wages and drives competition for relatively modest paying jobs through the roof.

If every job market in the US worked like the Academic job market, the country would be hell on earth.

> It is convenient for the country to attract and retain their talent, of which there is currently a shortage.

I think this subtle misdirection is how the lie gets slipped in. It's not that there is a shortage for talent. It's that there is a shortage for talent available at the price point that corporations want. Does that mean we should create an indentured servitude system to make that available?

The whole idea of artificially creating an immigration system expressly for those willing to get exploited is pretty distasteful to me and harkens back to a bunch of other distasteful ways this country used to procure its cheap labor. The bandaid has to be pulled off at some point. People will get hurt in the process. But if this isn't rectified now, even more suffering will occur in the future.

In my view, we can never not have a shortage of talent when it comes to scientific disciplines - at any pricepoint. The beauty of bringing more researchers to the country is that they eventually help create as many or more jobs than they "take".

What leads you to say that? I'd take the counterpoint. Much of what passes for "science" these days is glorified p-hacking for conference proceeding churn and sweatshop adjunct work to pump out a minimum amount of monetizable IP to underwrite an ever bloating university administration complex. It seems closer to scientism than to science for me. And I think that is the natural result of intellectual pursuit being diluted into a commodity.

On the comment regarding "p-hacking". I think that is a separate issue altogether and irrelevant to the discussion. Universities have essentially become businesses and they push scientists to "publish or perish". If anything, it is a case for democratizing science even more.

I think you meant to say there will never be an excess of talent? (If I’m reading your comment correctly)

Yes, sorry for the double-negative

>H1B visas are very important for foreign students in these fields particularly those hoping to pursue careers in academia.

I don't care. It's not the US's job to educate the world. Until someone helps me get into university again I don't give a rat's ass about foreign students coming in. Those students are typically in the higher earning income brackets for their countries.

Sorry. I was going to type something rude. The thing is, by "educating the world", the world educates us. We need to do a better job at helping students like you pay for college, but this does not mean that we close our doors to everyone else.

Student visas are essentially free money for the countries that provide them: unversities charge them more; the students tend to be rich and spend a lot; a lot of students immediately return home and don't actually take any local jobs. There's no downsides.

There are limited seats. At the extreme end you have medical school, where foreign students likely are taking away a position from a capable citizen.

Buddy, you gotta help yourself get into and pay for university.

Why should a university give a "rat's ass" about you when a foreign student will pay more?

Or when they can use foreign students as a ridiculously cheaper labor pool.

Exactly the rub then, isn't it? I got the token half off 200% markup while I watched my graduating HS class get nearly free rides. I had similar grades, but was mostly straight A (but in harder honors classes) and was white and male.

The issue probably wasn't that you were white and male.

The issue probably was your parents made enough to not qualify for extra financial aid.

That sucks, and I empathize. I'm living it trying to get finaid for my kids private school.

But, let's not blame immigrants for a broken system. If anything they aren't to blame because they didn't vote for it.

The issue is many things, but what I listed is appropriate. All of the women in my class with my GPA got more money than I did, and all of the minorities (as few as they were) did as well.

Two instances stand out to me: Someone 1/64th or less Native American getting scholarships for that, and a girl who just by accident ran into (literally) the rowing coach and was offered a scholarship on the spot.

If their projects are that important we should be spending more on them so they meet the minimum pay requirements.

Part of the problem is that in basic science, it is hard to measure the value of a project until years in the future. But we do allocate money on academic projects -- through federal grants.. nevertheless the base salary for a postdoc is pretty standard across the country, an subpar compared to industry. Hence, salary does not necessarily correlate with importance of the research.

Postdocs are not even on H1B the vast majority of the time, because it's considered academic training. It's possible to get on a H1B if everything has been exhausted, but never the preferred option.

I did my Ph.D. in Physics on an F visa, and H1B was my only option when I became a postdoc. Already on the old system, I was paid significantly more than my US citizen colleagues due to the minimum wage. I’m worried about what this rule change will do to my field, where talented postdocs are hard to find. It also never made sense to me to invest many $100k in foreign Ph.D. students to then make it so hard for them to repay part of this investment to the country.

The problem with how academic and science work is currently structured is that there is an oversupply of labor at basement-dwelling compensation. Post-docs work for decades before getting better opportunities, and people are cheap labor in industry labs for long periods of time.

Foreign workers are partially compensated for low pay with a potential visa to stay, while this extra motivation is not there for Americans.

Having a native pipeline of education to knowledge work is important for long term competitiveness and is hard rebuild. If the US falls upon even harder times the foreign pipeline will dry up.

This pipeline is already drying up due to covid.

No doubt that H-1B has brought many talented people to the US that have made enourmous contributions. But this has been out of random chance rather than design. The program needs calibration.

The H-1B visa program currently requires to have a "bachelor's degree" or equivalent. When you open the door to any "bachelor's degree or equivalent" from anywhere in the world, results are pretty random. You can get a freaking genius, or a person that cannot implement fizzbuzz, and anything in between.

I think the solution is to give priority to the most skilled applicants in the pool, in this way, the cheaper talent will be left behind and wage depression becomes harder to achieve.

I think we should be expanding the H1-B program, but it is already oversubscribed and if this can reduce applications to the point where the lottery is not needed or is much more certain, I think that would be an improvement.

Reply on news.ycombinator.com