Hacker News Reader

Comment by Ultimatt | Hacker News Reader

Plenty of research doesn't need millions to do. If you're a mathematician or almost any kind of theoretician or computational scientist you need a laptop maybe the odd bit of cloud time on a bigger blob of compute, in the case of mathematicians you might need chalk. The millions you do need to spend though are on private per-paper viewing of literature, thats the real mind killer. If universities just had alumni access to literature they would rapidly just become subscription library services for theory people! I can work like 3 months a year and finance literally the rest of my time on theoretical work. But only if I don't have to pay £60 to read enough of a paper to realise its a worthless piece of crap.

View on HN

I should add those 3 months would also pay me about 2x better than a postdoc working fulltime in a university. Thats the level of pay disparity with industry, and the level of pointlessness of a university as an institution if you dont need a wet lab to work. A genuinely good business idea would be a virtual institution with library services.

You're forgetting the cost of funding students, but also there are costs in securing prestige for your institution. The university benefits when your grants are large and plentiful, so they encourage research faculty to pursue those projects which earn more money, even if it's not exactly what they care most about.

> You're forgetting the cost of funding students, but also there are costs in securing prestige for your institution.

But why does he need students to do his own research?

I made a similar calculation and went into industry. I am now at the point where I could self-fund an indefinite post-doc (with benefits) at the risk-free rate of return. By the time I would be going up for Full Professor, I'll be able to self-fund an equivalent salary -- again, at the risk-free rate. And I don't have to worry about the inevitable layoffs coming to higher-ed, tenure be damned. And I can do my Science from a ski slope or a beach instead of from some random rural town.

If you want to do Science, just spend 1-2 decades in industry and save your money. If you want to lead "The $LASTNAME Lab" at "Famous University" where you have "X PhD students" and "Y post-docs", then you're going to pay for the silly prestige game by giving up actually doing Science for the rest of your life.

The situation may be different for people who need wet labs, but in Math and CS here's just no reason to bust your ass to get tenure. Just bust your ass in industry instead. You'll be working for The Man instead of doing what you really want either way, I promise, and honestly industry often has more interesting problems than whatever happens to be of interest to whichever person was willing to live in an expensive boring place on a small salary while shoving their ideas into the Heilmeier Catechism.

> You'll be working for The Man instead of doing what you really want either way, I promise

As an academic who does work on my own projects all the time, I'll have to disagree. I mean, yeah I have to do things I'm not thrilled about, like advising students and filling out course evaluations. But at the same time, I only work 4 days a week, 28 weeks of the year, and those days I'm done at noon. The rest of my time I spend working on my projects, which are 100% what I'd be doing if I were to come across a billion dollars tomorrow.

> If you want to do Science, just spend 1-2 decades in industry and save your money.

I see this advice a lot, and it's logical. But I don't think practical. If it were, I think we'd see more people going this route, but I'm not aware of too many (well, any really). Can you point to anyone who has amassed a small fortune in industry and now is a successful independent researcher?

It seems like the more popular route is to spend 1-2 decades amassing wealth and then just retire on that money. Or to invest it. You'd be what, 40+ at that time? That's a prime age for family rearing. The kids' college funds need to be swelling at that point. At such a critical time in a family, who in their right mind would leave a lucrative career to be an independent researcher, a job path that is largely reserved for independently wealthy people.

If you want to blow your family's nest egg at 40 by pursuing an independent research agenda, go right ahead, but I think that's a risk most people won't take. Even if successful, it's not going to have a positive ROI (that's not what research is about). That's why researchers spend other peoples money and not their own.

> But at the same time, I only work 4 days a week, 28 weeks of the year, and those days I'm done at noon.

TBH I have no idea what number of "bullshit hours" I work, but I spend the vast majority of my time working on things I'd probably be doing anyways. I certainly think what I'm working on is more interesting than what I'd almost certainly be doing if I had accepting one of my TT offers.

> If it were, I think we'd see more people going this route, but I'm not aware of too many (well, any really). Can you point to anyone who has amassed a small fortune in industry and now is a successful independent researcher?

This largely depends on how you define success.

They certainly aren't doing things like putting together performance bonus packets or playing the peer review game, so if you look to things like bonuses/promotions or journal papers in prestige venues they're probably not "successful". But isn't the goal to get off the bullshit ladder? You don't really think the prestige publishing game is anything other than BS make-work, do you? ;-)

> It seems like the more popular route is to spend 1-2 decades amassing wealth and then just retire on that money.

Sure. If your version of "what i would do with a billion dollars" involves lots of slopes and beaches then why not? What's wrong with this?

> Or to invest it. You'd be what, 40+ at that time? That's a prime age for family rearing. The kids' college funds need to be swelling at that point.

I know a lot of folks from my PhD cohort who went this route and took Professor of Practice positions for their kids' college years.

> If you want to blow your family's nest egg at 40

Who said anything about blowing? The principal should continue growing, just a bit slower since you're spending some of the interest...

> At such a critical time in a family, who in their right mind would leave a lucrative career to be an independent researcher, a job path that is largely reserved for independently wealthy people.

A STEM PhD, after 10-20 years in tech or finance, IS an independently wealthy person... that's the whole point.

> but I spend the vast majority of my time working on things I'd probably be doing anyways

But is it your project? Or is it your employers, and do they actually own it? If you’re fired, do you still get to work on that project?

> what I'm working on is more interesting than what I'd almost certainly be doing if I had accepting one of my TT offers

I’m curious, why wouldn’t you just decide to work on a project that was more interesting to you?

> But isn't the goal to get off the bullshit ladder?

The goal would be to do research. So my question is: who is doing such independent work? Who are these researchers and how are they disseminating their work? If there aren’t very many of these people, can we not conclude this mode of research is not a viable alternative to mainstream research?

> What's wrong with this?

Nothing is wrong with this. But the suggestion is that if someone wants to do research, they should amass wealth in industry and then do research. My point is that close to 0 people end up taking this route (at least from where I’m sitting, maybe there is a whole community of multimillionaire tech researchers I don’t know about).

So I wonder how viable this path is? Do people just not consider it? Or do they lose interest in research over those two decades? Does life get in the way?

I guess my overall point is that, most people who want to do research, just do it. Spending decades amassing resources just to get started seems backwards.

> took Professor of Practice positions for their kids' college years.

And so you’re saying they funded their own research with their wealth at that point?

> you're spending some of the interest...

I thought we were talking about funding a research agenda. From where I’m sitting, that costs millions. If you’re making millions in interest income, I think we’re talking about at different scales here.

> IS an independently wealthy person... that's the whole point.

My meager research agenda costs roughly $1 million per year (no special equipment or facilities). How long can a retired tech worker burn cash at that rate and still make it through retirement?

Maybe the math works out but I don’t know.

> But is it your project? Or is it your employers, and do they actually own it? If you’re fired, do you still get to work on that project?

Varies. I optimize my employment situation for my personal ROI.

But it turns out that making money off of even excellent scientific work is often very hard and involves insane amounts of slog. Like 1 hour of science to 10 hours of not-science. So usually employment and assignment has better risk-adjusted return. Not always, but usually.

> So my question is: who is doing such independent work? Who are these researchers and how are they disseminating their work? If there aren’t very many of these people, can we not conclude this mode of research is not a viable alternative to mainstream research?

I know there are several people like this in my field. They publish a paper every year or three, often in smaller venues. Or just throw things on arxiv.

> So I wonder how viable this path is? Do people just not consider it? Or do they lose interest in research over those two decades? Does life get in the way?

I think the middle and latter options happen quite a lot. I meet a lot of early retirees on Tuesday mid-morning lifts. I think a lot of people lose their passion for research in their 30s. This happens in the academy as well. Sometimes after amassing wealth you realize "hey, you know what I enjoy more than doing world-class mathematics? Chilling in my cabin with my family and skiing."

> And so you’re saying they funded their own research with their wealth at that point?

They took the PoP positions to pay for their kids' college without spending down savings. A sort of middle ground between working and retirement, and also a good way to build some collaboration opportunities.

> I thought we were talking about funding a research agenda. From where I’m sitting, that costs millions.

To be clear, the premise that started this thread is the "a guy and his laptop/occasional cloud spend" style of research that's prevalent in CS and Math. Yeah, if you need $1M/yr then you need an income stream or a different level of wealth.

> I guess my overall point is that, most people who want to do research, just do it. Spending decades amassing resources just to get started seems backwards.

I think this is reasonable advice for people who (1) need something more than a laptop and modest cloud compute, and (2) are absolutely sure that they would rather pursue that particular research as opposed to retiring early, and (3) are certain they they will be able to retain funding streams for the duration of their career.

How about those world class mathematicians working at subway?

What about them? There’s no way they would be able to get an academic position, which requires orders of magnitude more mental health fortitude than an industry position.

Look up Zhang.

One big thing is it's basically impossible to start doing research when you are 40 or 50 after spending 10 to 20 years not doing research. You might imagine that you can, and I do know a few people who think they can, but no it's extremely difficult and basically not happening. Is there literally any highly cited journal paper with a first author who didn't do research for 20 years?

Also, you have way more energy, drive, creativity when you are young and you just spent 10 to 20 years on other things instead of doing research during a time when you could have had the most impact.

> A STEM PhD, after 10-20 years in tech or finance, IS an independently wealthy person... that's the whole point.

Was an independently wealthy person, as the US job market was rigged to favor upper middle class professionals. Nobody knows if that's going to be the case for the fresh PhDs entering the job market today. Those lucrative careers may be threatened by any number of reasons, including political pressure and/or instability, more PhDs entering the industry, increasing competition due to remote work, and AI.

This is a good point. I think there are still opportunities, and will be for the next 10 years, at least for good CS PhDs. Certainly not for all PhDs in all disciplines.

> rigged to favor upper middle class professionals... remote work

In the case of PhDs, it's exactly the opposite. Those jobs have very weak immigration moats compared to eg law, medicine, engineering, even software devekopment. The vast majority of phd students in CS are already non-residents and that's been true for decades at this point.

For the same reason, PhD labor is probably the least effected by remote work.

> But why does he need students to do his own research?

That's a phenomenally good question. Short answer: because of incentives in academia().

To be high up in the academic hierarchy(), you want top-ranked publications. So you work on those and score a few - good on you! However, a researcher in the same field outdid you by publishing their own work AND supervising a student. That increased their output, so they get the promotion. Now they get more means to attract phd students and postdocs, all collaborating with them, further increasing output.

The nasty thing is that this cycle has knock-on effects. Do everything by yourself and you're stagnating; collaborate and you are rewarded with more recognition, promotion, more citations, more reviewer invitations, more invites to give talks/keynotes, etc. And: more quantity and quality in those trigger more and higher-valued of those. Suddenly you're growing into a "respected member of the community" - which strengthens your position at your instituting. Stronger position locally is more local options for growth, leading to the next cycle.

TL;DR: Because the direction of academia is in the hands of managers (uni admins, politicians, grant body admins), and managers need quantifiable data... so you're incentiviced to make your numbers go up.

I totally agree. Academia is a prestige game. If you just want to do science, you're better off taking a non-traditional path.

Benefit to that is you are training the next generation. Do research on your own, good, but have ten grad students working for you and you can basically multiply your output by ten. Note, the grad students a majority of the time won't be able to do the research without your advice, this is just a fact; there are exceptions but that's just how it is the majority of the time.

Also, who is going to defend your scientific legacy after you pass on to the great ivory tower in the sky?

> But why does he need students to do his own research?

What a... weird way to phrase this.

"Need"? All the faculty I've worked with have wanted students to work with them. Science is frequently a collaborative effort, and having more than one person working on a project is often a good way to make the project better. They generally do not see students as a burden, but as an investment into the future of the field (and their own success, of course, but in a collaborative sense).

It's fine if you decided academia wasn't for you, and I don't doubt even a little bit that people do exist who see students as a burden to be dealt with. But my own experience has been vastly different from that, so it strikes me as odd that you simply assumed a cynical stance off the bat.

Anyway, your general stance seems very cynical. I like doing my own work, but I also like collaborating with people, and I know I'm not alone in that.

Cannot you just get a library card at the university? I live next to a university where I have no affiliation, and if I pay $150 a year I can eat all I want. (Though I do think I have to be onsite as there is no remote access.)

Another trick is to sign up for the adult ed classes, or basically retiree classes to get access. I've even heard of people doing this to get cheap health insurance!!!

It depends on the field, but I don't think access to papers is such a big deal anymore. Preprints and open access mandates have become popular enough that relevant papers are usually available for free through legitimate channels. When I'm working from home, I rarely bother starting a VPN to access a paper, because it's usually not worth the effort.

A bigger issue that affects independent researchers (and academics who are the only ones working on some topic in their department) is the social context of doing research. Even if you are a big name in your field, your ideas are probably not that special, and neither is your ability to execute them. For most researchers, the real value comes from having regular discussions with other experts.

There is, of course, sci-hub, if your morality aligns with that.

Do you not have access to Sci-Hub somehow? Or have I misunderstood the problem here?

Sci-hub has become significantly less useful because all new uploads are paused pending resolution of a court case in India. That includes anything published in the last year. A large portion of a working academic's reading load will be keeping up with recently published literature.

But for papers published recently you can email an author and they will send you a PDF.

Just imagine how useful search engines would be if you had to email someone every time you wanted to click a link. Just keeping up with the literature is typically a dozen or more papers a week, plus skimming dozens to hundreds more if you're actively doing research. That's a lot of email and it's a lot of unnecessary burden for the very nice people who reply.

Piracy and open access are the only scalable alternatives to institutional access.

Reply on news.ycombinator.com