The other side of "academic freedom" | Hacker News Reader

The other side of "academic freedom"

matt-welsh.blogspot.com

124 pointsphenylene56 comments

I really wanted to get into academia because I had the perception that you got paid for thinking smart thoughts and occasionally writing about them, but when I actually tried academic research it seemed to be about 80% politics/grantwriting and 20% research delivery, so I went into private industry. If you have sufficient reputation to have a sinecure from Google, that's certainly an option for subsidizing research. Even if you don't, a lot of research in our field is not terribly expensive to do, and the materials/collaboration required to do it are within the grasp of e.g. anybody who runs a business which gives them sufficient scheduling flexibility. You can even get published in Real Academic Journals (TM), which surprised me.

Certainly, during a lifetime in academia, I have observed that the most successful professors are not the ones who work on what they like (with very few outstanding exceptions) but instead are those who learn to like what they must work on. Where 'must work on' is defined pretty much by what is described in this well informed article. The funding trends and the ability to anticipate them so as to be amongst the first in the line being the most important. Knowing the right people certainly not to be neglected. Ability to do any kind of useful or worthwhile science a long way down the list.

This undesirable trend has been lately exacerbated by the race for the funding and the attendant unhealthy over competitiveness. Academia is now a very different world to what it used to be, where a fellow would get enough ad-hominem funding, on the strength of his proven ability and membership of an elite college. Indeed, without that, the phrase 'academic freedom' has lost most of its meaning.

I believe that 'academic whoredom' would much more fittingly describe the current state of affairs.

As someone who relatively recently decided to go into industry instead of academia after finishing my PhD, my reasoning was almost identical. Industry seems like a better chance to work on interesting problems and have a bigger impact.

However, I'm really regretting it at the moment.

I thought I'd at least have a say in how I approach problems, even if I didn't get to choose the problems I work on. Instead, I've literally been told, "Your job is not to do science. Your job is not think. Your job is to click a mouse. Do exactly what you're told in exactly the way you're told to do it. Now, stop asking questions and stop trying to think for yourself."

I realize that the people I'm working with have a _lot_ more experience than I do, but I do think there's value in considering alternate approaches to solving problems. A lot of the fault lies with me, too. In the case above, I didn't clearly communicate the business impact of what I was suggesting to my mentor. Either way, it's frustrating.

Most of it is finding the right team to be on. I still think I chose the right company, I just need to find a way to be on an R&D team. I'm under a 2 year contract, so I can't leave even if I did get frustrated enough to.

At any rate, know what you're getting yourself into. I did two internships on similar teams, but at different companies. Things are definitely done differently here.

There is certainly of lot of interesting work in industry, but it's very easy to wind up stuck doing menial repetitive tasks if you're not careful.

That's unfortunate. I'd be curious to know where you ended up taking a job. I have been surprised at how "academically inclined" Google is; however, I am sure there are groups here that have less interest in research-type endeavors than my group. Industry is by no means homogenous in this regard, which is why I decided to save that discussion for a later post.

I'm in the energy industry, rather than in software directly. (I'm a geophysist, rather than in CS.) I'm at one of the majors, but it's probably best if I don't say which one.

There are certainly "academically inclined" groups where I work. I just need to work towards one of them.

I've enjoyed all of your posts, by the way!

This varying a lot by field is a good point. It also depends on what you mean by academically inclined: are you interested in doing research, in publishing the results of your research, or both? In chemical engineering, for example (not my own area, but some of my family are in the field), it's much easier to find a research-oriented industry job than to find a research-oriented industry job that will let you publish openly.

I'm a late stage phd student contracting in industry and finishing my theiss part time. I was chatting with a prof the other day, saying how I was pretty happy with my work situation. He said "so they pay you well, give you interesting work, and leave you alone to get on with it. Sounds pretty sweet to me". So it depends on the gig. I am particularly lucky, not due to good skill as a programmer, but due to good, relevant, people and project management skills.

If you're not exaggerating very much, then your employers are fools. Clicking a mouse could be done by a teenager on minimum wage instead of paying to hire a professional. Start quietly looking around for more clueful employers.

But what do you mean you can't leave? Indentured servitude isn't supposed to be legal nowadays. What happens if you leave anyway?

Your mentor said that to you? Nobody I'd call a mentor would, except in special circumstances.

I hope your circumstances get better.

Amusingly, I think his dig at the programming language community actually underscores something that I really love about academia and that I think is virtually unique: you can focus on things that are not immediately practical just because they're elegant or beautiful. It's awesome. I couldn't imagine much room for it in industry (although it's not entirely impossible).

Unfortunately, it also seems surprisingly rare even in academic CS. Certainly nobody I know at Berkeley operates this way :(. Maybe it's better in the mathematics department. For me, that would be the main draw to academia: being able to not worry about short-term usefulness and not needing to worry about marketing my work to the average programmer.

Besides, those "esoteric abstractions"? They solve real problems. In surprisingly simple ways. Better than existing solutions. The only reason that "real software developers" don't benefit is that they aren't willing to learn them. And it is not--and should not be--the responsibility of the PL theory community to get the average programmer up to date. And, even so, some people are really trying. It brings to mind the usual saying about horses and water.

Also, if I may make an observation: it seems his thoughts on PL design broadly reflect the philosophy of Google in general. This is just reinforced by the designs of Golang and Dart, as well as the promotional material surrounding the languages (presentations and the like). I personally think that this almost borders on anti-intellectualism: they seem to imply that something created by ivory-tower academics with an understanding of math cannot possibly be useful in the real world; only real software engineers™ make practical tools.

And that is probably the main reason that I would choose a company like Jane Street over Google any day :P.

Yeah, I'm probably more annoyed by this than it's worth. It really helps to lay my thoughts down on paper, and I still haven't started that blog I keep on considering. So HN will have to be my confidant for now.

As an academic/industrial researcher* in the programming languages community, I agree with Matt. And who do you think makes up a vast majority of Google's PhDs anyways? That's right, PL researchers who want to do do real work and solve real problems (e.g. Jeff Dean).

> The only reason that "real software developers" don't benefit is that they aren't willing to learn them.

This is a stupid response: its like saying, "users are just stupid, fix the users, not the device!" We are simply not willing to listen to what their real problems are. We aren't solving their problems, so why should real developers care about us?

> And it is not--and should not be--the responsibility of the PL theory community to get the average programmer up to date.

Does the PL theory community actually cared about programmers? There isn't much empathy on both sides.

> Also, if I may make an observation: it seems his thoughts on PL design broadly reflect the philosophy of Google in general.

If you are going to attack something on hackernews, please give at least a few details.

> I personally think that this almost borders on anti-intellectualism: they seem to imply that something created by ivory-tower academics with an understanding of math cannot possibly be useful in the real world; only real software engineers™ make practical tools.

Why is this so weird? You know, the ivory tower's output is super cutting edge research + people. They aren't supposed to be making practical tools.

* disclaimer: I don't work for Google, but I do work for Microsoft.

> The only reason that "real software developers" don't benefit is that they aren't willing to learn them.

Willingness is probably not the problem. Can you provide an example of such an "esoteric abstraction"? I am pretty sure, you will get an extensive list of downsides in the comments.

Monads are a pretty general encapsulation/sequencing abstraction being useful for more than just IO handling in pure languages. Although to be fair, they are starting to show up in more mainstream languages (ie Clojure)

A Monad is just a theoretic abstraction, not a useable one. Only specific Monad instances (IO,Maybe,List,etc) can be "useable", hence "useful". There is no generic "bind" for example, it has to be implemented specifically for every Monad.

While Monads are pretty useful in pure languages, they are often over-engineering in imperative languages.

What I consider the actual point that the OO world is currently learning from the FP world is abstracting over computation. OO can abstract data and state, but computation is rarely modeled as an object. Monads are simply a pattern you can observe often, when computation is abstracted.

Scala has monads aplenty.

I realize it's in quotes, but very few academics hearing the phrase "academic freedom" would think of "the ability to work on anything you like." They would, rather, think of what Wikipedia aptly describes as:

". . . the belief that the freedom of inquiry by faculty members is essential to the mission of the academy as well as the principles of academia, and that scholars should have freedom to teach or communicate ideas or facts (including those that are inconvenient to external political groups or to authorities) without being targeted for repression, job loss, or imprisonment."

I've been in academia for (including graduate school) a little over twenty years, and I've been astounded at the ways in which freedom of inquiry is threatened from without and from within. And while few American and European academics face imprisonment for their views or objects of study, many academics in other parts of the world do. And U.S. and European academics certainly can face job loss or sanction for being interested in "inconvenient" things.

I would encourage the author to confine his use of this phrase to its traditional meaning, lest anyone forget the far more important problem to which it refers.

But this is all wrong. "Academic Freedom" means that you are free to publish your findings in an objective way without having to please your investors.

Of course you have to do things that are interesting and impactful and the funding is less now than it has been in the decade prior, but that is much different than saying that academic freedom means doing whatever you want. I can just imagine Michelangelo writing this blog during his agony of painting the Sistine Chapel.

Two separate points:

1. I'm surprised that Welsh argues that tenured CS profs still have to follow what publication venues want to see. Why not just say, "Fuck it?", publish on blogs / arvix.com/, and let the field catch up to them? Certainly that's not a route to immediate promotions or status within the field, but there may be strong long-term returns to individuals who go this way and are vindicated over time.

(This obviously doesn't apply to non-tenured faculty or grad students. I'll also note that this point is a related observation, not a criticism of his argument.)

2. This stands out to me:

The final (and arguably most important) aspect of being successful as a faculty member is being able to solve new problems better than anyone else in your area. It is not usually enough to simply do a better job solving the same problem as someone else -- you need to have a new idea, a new spin, a new approach -- or work on a different problem.

Genuinely new ideas are actually quite rare. Sometimes the difference between a "new" idea and someone else's discovery or implementation of that idea can be just a couple months difference! (See Steven Berlin Johnson's Where Good Ideas Come From for one popular description of this.) Yet one person or group gets 99% of the credit / tenure / etc.

I'm surprised that Welsh argues that tenured CS profs still have to follow what publication venues want to see. Why not just say, "Fuck it?", publish on blogs / arvix.com/, and let the field catch up to them?

Because funding agencies don't like it. They'll look back at previous funding you've been given, and rightly or wrongly, look at the peer-reviewed articles that came out of it. If you're pushing everything to arXiv, then it looks like something went badly wrong and you can't be trusted with future money.

As the OP says, the problem is not just ideas, but how those ideas are disseminated and what that means to your funding stream. Saying "fuck it" is a really good way of screwing up your funding stream for years.

ad 1: Please also do keep in mind that there are career steps after obtaining tenure. If you say "Fuck it" right after getting tenure, you are almost certain to never make it thus far.

I also do agree with the other comment regarding funding agencies. Another problematic way that NSF does business (inviting professors for peer review that is) is that this virtually guarantees that some of your peers know exactly what you're doing, which reduces effectiveness of double-blind submissions substantially (to the point where it is hard to believe it works at all; didn't it ever occur strange to anyone that the same people from the same top schools are consistently successful? [with grants and publications in the top venues])

On the first part, that's true, but you don't have to actually care about those further career steps. Tenure you have to care about, because you're out of a job if you don't get it. But a tenured professor in CS has a permanent position with a good salary, regardless of whether they ever get another promotion or not. Some people do really care about the Full Professor title, but it's not mandatory that you take those further career steps particularly seriously.

A bigger issue post-tenure is the money side of things. Do you need significant funding to carry out the kind of research you want to do, e.g. because it needs reasonably expensive equipment or employee/minion labor? If yes, you do have to care what people with money think, whether it's funding agencies or corporate donors, and that constrains the research you can do. If no, e.g. because you work on your own projects and don't particularly need equipment or a large lab of minions (common in areas like theory and logic), then you don't really have to care about the funding agencies, either.

Thanks for the comments; I replied on the blog post itself. TL;DR: Even tenured profs care about helping their students get good jobs, which means publishing in top venues.

It depends on the kind of research you do, but one viable approach in some areas (which I've seen work in practice) is just to not have as many students. Plenty of tenured theory profs out there have one, maybe two PhD students at a time. They do still need to work with that student to send the student's papers to top venues so he/she can eventually get a job, but it's a lot less of a constraint than if it's a systems-style lab full of 5 or 10 PhD students who all need to be pumping out top-tier publications, and leaves considerable time to work on one's own projects with fewer constraints.

For 1, I can think of one compelling reason not to do that: It's harder to convince your funding agencies that your work is having real impact if you don't/can't convince the best venues in your field to publish your work. This is problematic because in your funding agency's eyes, it's impossible to distinguish between "I choose not to publish in the best journals" and "I'm just covering for my failures by claiming to only publish in arxiv."

A professor without any funding is a professor without a hope of getting tenure.

> My team at Google has a pretty broad mandate which gives us a fair bit of freedom.

The other points might be fair but this would have to be the exception rather than the rule in private industry and even Google I would image would never give enough leeway to someone to "monitor volcanoes with sensor networks"

The sensor networks research was work I did at Harvard prior to joining Google. I don't think Google is very interested in my deploying sensor networks on volcanoes, although (to be fair) I have not asked.

google.org might be, if there's a humanitarian spin on it.

Interesting example, as I actually know people at a company that have that kind of leeway. And are actually doing it. :)

I want to study astronomy. After first year I found it to be a dead end. So I went to private sector, got nice 9-5 job and hack astronomical software over evenings. More freedom, more money and my software is still used by thousands of astronomers.

I'm surprised that we don't see more in the way of "open source" computer science research, done by unpaid volunteer enthusiasts.

It may be difficult/expensive to get to research conferences without institutional support.

It probably is, but I wonder how much one needs time at research conferences to produce interesting work.

The word "wonder" is very deliberate here: I genuinely have no idea. In English lit, my own field, access to a research library (with peer-reviewed articles and books) is far more important than conference attendance, for example.

I'm not precisely sure myself - in CS, though, a lot of the research is digitized, and much of it is available for free online (if you know where to look - Google Scholar really helps), but as far as making useful research contributions goes, being able to provide citeable papers helps, and most of those are published at conferences.

"80 percent of success is just showing up" — Woody Allen

Use conferences to make connections and get inspirations and recommendations from talking to other researchers in your field.

Being present + papers > papers.

We have the Internet now. Conferences are more about career networking than intellectual networking.

I've actually found conferences quite useful for intellectual networking, although I'm not really an outgoing person, and it's not my personally preferred approach. I'd rather have some robust mailing-list communities, but they don't exist in all areas, and many people pay relatively scant attention to mailing lists. I do do most follow-up and collaboration over the internet, but in a number of cases with people I met at conferences.

In particular, I've found showing up at conferences to be a way to get your research onto people's radar who didn't previously know about it. Just by being there, and especially if you also give a talk, it can make people think of you as part of the X Research Field, and remember that you did some research on Y. In turn they're then more likely to engage with it, cite it, mention it to others when it's relevant, and maybe keep you in mind as someone to potentially collaborate with (if research interests align). All that could happen completely passively, by putting your paper on the internet and hoping people find it in a Google Scholar keyword search. But that's relatively low-probability. You could also try to approach people online, but "cold-emailing" people about research often ends up neglected, since people get a lot of email. Conferences by comparison give a setting where people are expecting to hear some talks and exchange short "what I'm working on" pitches over coffee or dinner.

I think it might even be true outside academia. Some startup/tech events are very focused on business/career networking, but others, like SuperHappyDevHouse, are great for intellectual networking.

It's not just the conference but also how you're networking and with who. I've had fantastic tangential intellectual discussions at tech conferences.

CS is a bit different in that conference proceedings are where a lot of the peer-reviewed papers are published. E.g. in graphics, SIGGRAPH is the top place to publish, and in HCI, the same goes for CHI. So attending conferences to present papers is part of the process of getting papers published. Journals do exist, but tend to be used for big archival papers: you might take 3 or 4 years' of conference papers on a project and wrap them up into a giant 40-page journal article for posterity. But they aren't really where recent research is being published & read.

Admittedly that's less true in some areas, where journals do have a larger role. For example, in machine learning, while it's common to publish at conferences like ICML or NIPS, it's also perfectly fine to skip them and just submit to JMLR (http://jmlr.csail.mit.edu/).

Journals do exist, but tend to be used for big archival papers: you might take 3 or 4 years' of conference papers on a project and wrap them up into a giant 40-page journal article for posterity.

That's the polite interpretation, at least. I'd say the pattern I see more often is that the exact same research gets presented repeatedly at 3 or 4 conferences before getting published as a journal paper. Also, the research was probably done by 1-3 people, but an entire group of 6-10 people will have their names on all the papers...

Oleg Kisleyov is amazing at this. He works for NOAA or something and is a leading PL researcher in what I believe is not his main job.

Also, Satoshi :-) and Bram Cohen...

I had been thinking along similar lines in physics research as a post-doc, and wrote my experience and ideas up[0].

Then I got fired (unrelated to me writing that post:), and went in a job interview with another prof, who have read my post, and started off the discussion with that. In the end, he have seen the value, the ideals, and that the problems outlined were valid (he also worked in the Bell Labs which is one of my inspiration).

In the end I decided that instead of joining another lab, or as the OP working in the industry, I'll wing it and see if I can set up things the way I want. So now my "main job" is to do research and set up the lab I would want to work in, and get the people who are just as ambitious on board. "Tall order" doesn't even begin to describe it, let's see what happens.

[0] http://gergely.imreh.net/blog/2013/02/academia-is-failing-bu... "The Academia is Failing But Not For Everyone"

This is exactly what I'd like to do with my life if I could afford to. How is it going so far?

Working on the foundations, as I still need a few months of Taiwanese visa to be able to start my own company here legally.

In the meantime:

- Figuring out useful products used in a physics lab that I can use for my research and sell it as well, plus other value added things (like software development, consulting) that can support the things.

- Reviving and expanding my academic network (have researcher friends in England, India, Bangladesh, Korea, Japan, Singapore, Hong Kong, France) and see what would fit them well too

- Expanding Taiwanese networks for actual researchers, developers, electronics, manufacturing, via the Geek Dinner I organized and our newly founded Taipei Hackerspace, as well as friends working in the industry

- Expanding my knowledge via the likes of $100 Startup, Lean Startup, The Art of Innovation, Losing My Virginity, ....

- Moonlighting with a local creative firm to learn more about real business and money side, and get more professional services contacts

- Setting up a website and prototypes in the next month or so (sooner rather than later).

If you ever wanna chat, I'm on gmail.com with the same username :)

That sounds tough but feasible. I think the consultancy bootstrap is a good idea. I'll get in touch via email, I'm quite interested in knowing how it goes :)

"How should we make it attractive for them [young people] to spend 5,6,7 years in our field, be satisfied, learn about excitement, but finally be qualified to find other possibilities?" -- H. Schopper

Interesting to hear the other side of this. I'm a PhD student in Computer Science right now, and a lot of the appeal of the job is the ability to get paid for working on projects that interest me. That said, I have a nice government fellowship for my PhD and a generous funding package from my university - if I was missing one or both, things would be harder.

Do you mind me asking where you're doing your PhD?

University of Waterloo. They've got some great research groups, the funding's good, and you get to keep any IP you develop over the course of your studies.

What are you working on for your PhD?

Well, it's early days yet (I'm just finishing up my coursework, and haven't gotten far into my research yet), but my proposal is to build generic collections data structures for general-purpose GPU programming (and related architectures). Single-threaded performance is hitting some pretty serious walls so parallel execution seems to be the best way forward, but memory layout can make orders-of-magnitude difference in performance and is hard to get right. My goal is to build something analogous to the STL to make it easier for the average programmer to program for GPU.

Do you know about the 'thrust' library?

http://thrust.github.io

This is one of my main worries about my impending graduation. Right now, I work on what I find interesting, publish when I have what I think are good results, and have not found that anyone has tried to dictate what I should be working on. Once I graduate though, I'll lose that luxury. Unfortunately I don't know of any "real-world" job that provides such freedom; perhaps grad students are only afforded it because we're so affordable.

Right. It is rather easy for someone with a CS eduction to earn ~$40k a year and spend most of the year in freedom. Just get a contracting/freelance gig. The hard person is getting $100k or $200k or more and still having freedom.

I have a question for tenured profs, why one needs to apply for grants after getting tenured?

my guesses: to increase its own salary; to buy equipment; to support the group.

is it possible, to do research alone (maybe having one or two grad students) ?

Those are basically the reasons you'd need grants, yes. And yes, it's possible to go another route, and do research alone or with few students, in which case you don't need many grants (perhaps any). In areas where not much grant money is available in the first place (pure mathematics, say), that is what people do by necessity.

The author of this post comes from an area, systems, where large-budget, many-student labs are the norm, partly because of the type of research, and partly just because the (relative to other areas) availability of funds means most of your colleagues are going the large-budget, many-student-lab route.

The university takes an overhead charge on any grant money, perhaps 40% of the total. The professor can then use the rest of the money to purchase equipment (computers, tape recorders paper), pay research assistants (grad students), travel to conferences, buy datasets or software like SPSS, or pay themselves summer salary or extra salary.

If your research requires none of these things, and you have tenure and don't care about showing your field that you are capable of bringing in grant money to help fund your department, then you don't need to worry about grant money. This is probably less than 1% of the tenured professors.

It also has to do with ego; your peers will know who is bringing in the big grants and who is not.

Reply on news.ycombinator.com