Deep learning, Yann LeCun, Yoshua Bengio & Geoffrey Hinton,
Nature volume 521, pages 436–444 (28 May 2015)
Strategy: publish in the high-impact established Nature-family journal, but publicly boycott the no-impact new Nature-family journal. Win-win.
edit: added -family suffix to clarify that I completely understand these are two different journals from the same publisher.
edit 2: I now understand why the editor of the new journal was hitting senior people up for personal meetings at ICRA last week. He didn't want to hear about our work: he wanted hints to help him on positioning.
there's a huge difference between publishing one really general public descriptive article for the whole field and stifling research by having actual research papers published in a closed journal.
Their policy: "Nature Research journals support posting of primary research manuscripts on community preprint servers such as arXiv and bioRxiv. Preprint posting is not considered prior publication and will not jeopardize consideration at Nature Research journals."
Standard practice in the modern AI community is preprint to arXiv first, then out to a journal for review. All the papers are free online before they even go to Nature. So this policy does not stifle or hinder research.
They also do not take copyright - it stays with the authors - unlike other giants such as IEEE.
To my mind the legitimate argument is about publishers taking undeserved profits from unpaid labour while adding little value.
> To my mind the legitimate argument is about publishers taking undeserved profits from unpaid labour while adding little value.
That argument also comes down to hindering research. Funds that could otherwise be used for research are now used to buy subscriptions to this journal.
Nature journals actually do provide professional copyediting services for all articles accepted for publication. For commentaries (from my own experience), the editing process at Nature is quite intensive.
That just pushes out the argument by one level of indirection.
The publishers 'curate and edit' the board of editors for their journals. In turn, these boards of editors curate and edit the articles that go into the journal.
Nope, they don’t. The “volunteer” reviewers are drawn from past published authors, who in turn had their articles selected by previously published reviewers. Turtles all the way down.
All the conferences in my field have literal security guards checking the badges that are only handed out after paying hundreds of dollars to register.
Of course, but you don't need to attend the conference to read and benefit from the research, and that's the important thing.
The journal business model usually requires restricting access to the research, while the conference business model still works when the research is free.
They're boycotting the new Nature Machine Intelligence journal, not the old Nature journal. It's possible to be against a policy while still taking advantage of the policy for game theoretic reasons.
Not that I'm for journals, but those heavy hitters typically have their own brand in that of Google or Microsoft etc. Without an external 'brand' of Nature the importance of working for the recognised companies when submitting your research papers goes up.
That's what makes their support so important. If they're not publishing there, then the external brand value of this specific journal goes down, which means that even for less heavy hitters, the best "brand" value is still in the journals that actually make their work as widely available as possible.
People are forgetting what year we're living in, 1818.
It is not possible to publish something without setting moveable type made of lead. it would be different if there were an Internet or something but that's way in the future.
but there is substance to the joke above. so either its over people's head (as it seems to be on you) which is a shame as HN tends to have fairly intelligent culture, or else its not lost on the audience but they don't appreciate anything inexplicit or non literal, ie, humour, which is also a shame
The substance is that in 1818 you can't publish a serious journal to be read by scientists all over the world without expensive typesetting, printing, and so forth. (Setting moveable type, etc.) Typesetting a journal is hard work with high fixed and variable costs.
To give you a sense of the state of our industry, here is an example of an innovation just four years ago and its effect:
>1814 – First cylinder presses
>Friedrich Gottlob Koenig and Andreas Friedrich Bauer build their first cylinder press, which is much faster than the existing flatbed presses. One of the first customers is John Walter of The Times. The first issue of The Times that is printed with the new presses is published in 1814. The press is installed in secret to avoid sabotage by disgruntled pressmen operating the existing Stanhope presses. The machine is capable of printing over 1100 double-sided sheets per hour. In 1817 Koenig & Bauer return to Germany and start building presses in an abandoned monastery in Würzburg.
You can click that link for some pictures. This press needs to be paid for, as well as having ongoing operational costs that include a lot of meticulous manual labor (mentioned above) to set each and every page.
Here is another link which I rewrite to the present:
>There are many reasons to celebrate the advent of the steam press in 1814, as well as reasons to worry about it. Steam printing brings the cost of printing down, increases the number of possible impressions per day by four times, and, in a way, we might say that it helps “democratize” access to information. The Times proclaims that the introduction of steam is the “greatest improvement” to printing since its very invention. Further down that page, which itself is “taken off last night by a mechanical apparatus,” we read why the hand press printers might be concerned: “after the letters are placed by the compositors… little more remains for man to do, than to attend upon, and watch this unconscious agent in its operations.”
As you can see it talks about "the democratization" of information. Still, my point is that the fixed costs of this publishing model are still high, and that it makes a lot of sense to operate on this publishing model for this reason alone. After reviewing my links and the technology in question, I hope you can readily agree! In 1818 typesetting and printing is an expensive, hardware-based, industrial business. It makes sense to have a publishing house behind it.
Is there any part of what I've just written that you disagree with?
See, that contributed snark WITH substance. Gold star.
I’m not sure I totally agree with the parallels, which is probably where some of the initial confusion arose from. The printing press significantly reduced the costs of publication, allowing unprecedented access to the written word beyond the aristocracy. I don’t think anyone disagrees that Nature provides a valuable service and that it “makes sense to have a publishing house behind” peer reviewed literature. Rather, when the cost to produce peer-reviewed content can be taken on by tax-subsidized content production professionals, and demonstrably WITHOUT an older journal that extracts exorbitant fees without producing significant value beyond its brand-name, it makes sense to challenge the value of the brand.
If you have to pay royalties to use a digital font, it stands to reason you have to pay royalties to the keepers of information. Why that's elsevier et al all I don't know, but they are is the only logical conclusion.
Someone spent time and effort designing that font, and it's reasonable that they should be compensated for it. (Giving them the right to censor any use of the font so they can run a protection racket is a awful way of doing that, but that's not the point I was addressing.) Nature contributes nothing (today - they used to contibute printing presses) that couldn't be accomplished by publishing a first draft and soliciting reviews.
look at what they had to write for that gold star tho, to make the same point! good grief! ppl on HN be self employed, working few hours and all, but still, thats time he aint gettin back ;P
This forum in general values substantive comments like the second comment the OP made and doesn't value snarky one liners like the first one. That's the trade off for the value folks get from this forum. Other forums may have different cultures that may be more suitable for this particular viewpoint.
Everything? They publish basically extended abstracts, nothing that could be confidently reproduced, nor code ... or I just don't know how to read it and find supplemental material. There's a lot in the bibliography. Really just a journal article, an entry in a log book.
Kind of a hard stance to enforce. If you work for basically any major corporate entity you are probably sacrificing integrity to do so. So it seems you are calling the kettle black here.
Not quite every corp out there has a business model of specifically misusing and abusing personal information, manipulating peoples' emotions, etc., etc.
I work for a rather minor corporate entity anyway, does it mean that I am always off the hook? /s
Just because their parent company might not be morally in the clear, that doesn't mean that moral criticisms by its employees is automatically or even more likely to be invalid.
Oh, it might be valid, in regards to the magazine, alright. It just smacks of hypocrisy when their research is pretty much bound to be used for unethical purposes.
Jeff Dean is head of AI at Google. He's one of the most prolific computer scientists there, and has worked on many big projects such as MapReduce, BigTable, Spanner, GoogleBrain and TensorFlow.
Yann LeCun is head of AI at Facebook and kinda knows as the father of convolutional networks.
There's a bunch of other ones from other big companies or big ivy league universities.
Admittedly I work outside academia (though my wife is a tenured professor) and thus have some unavoidable naïveté about how the culture and community work, but I’m always confused about why universities aren’t just sponsoring the organizational costs of coordinating the review process, and establishing community-driven open journals to replace the rent-seeking corporate journals. What am I missing about the benefits of the status quo keep it so entrenched?
>What am I missing about the benefits of the status quo keep it so entrenched?
The entrenchment is due to perceived (and cultivated) prestige, less so about tangible benefits. Prestigious journals have good brand value, they spend money cultivating it and they extract earnings from it.
Researchers and administrators trade on journal brand value for career advancements. They benefit from a highly prestige-stratified journal catalog as a tool for differentiating themselves from their peers, it's much easier to justify a promotion based on 'X publications to prestigious journal A, B, C' than 'bunch of people in the community think this is a great result', especially to a non-expert.
I still think that "Journal of the MIT AI lab" or "Tokyo University Official Journal" would have no problem in establishing prestige if they were to try it.
I think you're probably right. However, I wonder what sort of retaliation the first institution to do this might face from publishers. The publishing industry will have a mighty incentive to kill it and they tend to play hardball.
Unless some sort of millionaire / billionaire (Thiel / Musk / Bezos, i.e. one of guys the hated by HN, just kidding) will be behind this open-source journal.
In all seriousness, I have always thought that Blockchain could pose a good solution for the issue of peer-reviewing scientific papers. It provides a descentralized infrastructure with "tokens" that can be transferred to provide "prestige" to papers and authors.
Well, I’m not sure about “prestige”; it seems like another arbitrary metric to be gamed.
However, I would like to easily see a non-political db where you can easily track the paper graph to figure out which papers are based on discredited research and/or researchers taking money from special interest groups.
Any (specific) idea how to get to that data? I might be interested in including this with some other, more general information-flow model{s,ing}. This part seems like making it publicly available would be a good thing, and possibly necessary, and likely aligned with the goals a likely choice of organization to host/maintain the technology would have.
i'm moderately hopeful that we'll get that sort of querying for free* as the semantic web matures, and we can better query federated linked data. university libraries are pretty keen on making that dream a reality (and this is the sort of use-case that makes them want to be able to do that, though they'd probably pitch the rosier version, where they show the indirect impact of a well-regarded paper or study, because that sells better than the doom-n-gloom version described here, even if that version is just as important).
* or if not "free", a very small marginal cost, in terms of say data aggregation and programming effort
...right, but then you have to figure out explicit metrics for determining what "prestige" is, and you have to explicitly implement the mechanics of it. i think this is a worthwhile research project of its own (or many research projects). but the mapping is very much non-trivial. i think the amount that people complain about how poorly the usual social media ontology ("friends", "likes", etc) maps to the "real"-world equivalent shows that this isn't something were you can just be like "we'll implement a first-pass of the blockchain thing now, and we'll refine the granularity of our prestige tokens later". thanks, but no thanks.
which is all to say, like so many of the things people want to use a block chain for, the hard part is the politics and the real-world nuance, not the immutable ledger. the immutable ledger is not trivial. but compared to the other stuff, it's not all that hard (and is certainly more solved, at this point).
You should cite "A Proposed Currency System for Academic Peer Review Payments Using the BlockChain Technology". Abstract here: http://www.mdpi.com/2304-6775/5/3/19
I think this is one of the rare cases where blockchains can actually provide good values: we could potentially get third parties of trust eventually to make a blockchain unnecessary but for now, the established parties are unethical and entrenched.
I hope this publication gets an implementation soon.
That works for experimental stuff, but a lot of work is theoretical.
Consider most of mathematics or theoretical physics. There is no real replication here. Instead, the test is strong peer review, combined with publication to a wider audience.
Similar issues occur for review articles, which are about analyzing data already present rather than analyzing new data. Moreover, this would also hurt actual replications. It kind of seems like being the last to replicate something would be without compensation.
The way to think about this is an article in Nature is like winning a tiny Nobel prize, with no prize money. You can make whatever other sensible publication outlet you like, but it's hard to create something that has the prestige of winning a tiny Nobel prize. That's Nature's value proposition. Brand, history and exclusivity.
This is why you can't get too angry at the scientists who continue to publish in it. If you kill yourself to figure something out worthy of a tiny Nobel prize, are you really not going to claim it? It's not like claiming that tiny Nobel prize is gonna hurt the dissemination of your work. The higher profile of your work will easily compensate for the fact people have to jump through a hoop to read it.
This is why I love sci-hub. It just destroys the hoop. It doesn't need to figure out how to replace the brand, history and exclusivity of publications like Nature. Along with government-mandated open source laws, it's one of the only realistic ways to actually disrupt this industry.
I don't think Sci-Hub breaks the loop. Researchers still need to publish with the mini Nobel prizes to advance their career in a highly competitive field, and governments/funders still don't want to dictate where researchers get to publish their work, so in effect, what happens is that governments mandate Open Access and pay the same old publishers significant (and disproportionate) amounts of money to make the single articles they fund openly available. In other words, the costs of research still don't go down, and the government-mandates were happening without Sci-Hub as well...
At the university I'm at, every time the library dropped its subscription to a particular journal there used to be a strong backlash from academics at the university.
Since sci-hub, the library has been able to drop a bunch of journals and for some mysterious reason none of the academics seem to give as much of a shit anymore. Obviously, no-one has no idea why..
That's basically how major journals got started, long ago. Most were essentially not-for-profit. But then, since the 70s, they've been acquired by large publishers. And those new owners have been monetizing those journals' prestige and market share. Given the inertia of the academic community, they've had a long run at that. But the end is coming, I think.
Just as with the established journals, organizing it isn't that influential and relevant compared to actual editorial board; and it would be likely or even expected that the board of a publication organized by some university would have just a couple members from that university and dozens of respected scientists from outside.
It's generally not considered conflict of interest, as the boards transcend single institutions and their members tend to be individuals in the stage of career where their individual reputations that matter more than the reputation of the institution they "represent".
Agreed. Within my own field, there are less than 10 people who are presently working on the same topics as I do. Consequently, we are constantly asked to review each other’s papers, and it is pretty trivial to tell who the author is, either because of the views and interests shown in the paper, or because of language use (i.e. you can tell that the author is a native speaker of the language in which the paper was written, or from their unidiomatic skills in it you can often guess at their native language).
The hard part in this is remaining silent and pretending you don’t know anything about this, when you meet your peers at conferences.
It is not blind for most science fields (I was not aware of any that were blind before reading this. PhD in Earth Science). What fields do you know of where journals send out papers for review with the author and his/her affiliation removed?
I've seen it exactly once (in a numerical analysis journal). Authors were blacked out. This was undermined almost immediately when the authors wrote in the introduction "... following previous work of the present authors [1]...".
Even if all these clues could be effectively removed, the nature of the research work, tools used, writing style, all unambiguously indicate authorship to someone in the field. I think for this reason double-blind review is not popular.
Which is not to say that it would be great if it were possible...
Whether peer review is blinded depends on the field. Some of the major ML conferences like NIPS are blinded, but in general most journals are not. The reviewers see the authors and affiliations and the editors see everything.
The problem is that changing an existing culture that is deep in everything how this system works is difficult. This goes way beyond the mere question of publication, it's still pretty common to hire people based on high impact factor journal publications etc.
You're faced with a system that's controlled by the people who made their career in this system. They're not the most qualified to improve the failures of that system. Changing it is possible, but it's hard and it's a slow process.
Well, the Open Access movement has been going at it for more than two decades now, so don't get your hopes up...
(That said, it's definitely not like nothing has changed! There's a lot changing at the moment, and it's not quite sure what the system is going to look like. Traditional publishers aren't looking too bad yet, though.)
Don't get me wrong, I think the current publishing model is archiac, exploitative, and needs to change, but this whole "taxpayer funded research should be freely available argument" seems to me to oversimplify the issue and ignore the facts.
Are you claiming that most research is already freely available and yet everyone is, for some unspecified reason, still paying publishers for an extraneous service?
The NIH site links right to PubMed Central from which it seems like it'd be 'easy' to find a paper for which I know the title. None of the pages on the NSF site ever seemed to be intended for someone that's not contributing to the 'public access repository'.
Would you demonstrate your claim and provide a link, for an article published in Nature, to the same freely available version that "must be made"?
It seems pretty fantastic that libraries would spend taxpayer money for access to journals containing papers that are already freely available. I'm very confused why that would be.
Can you give an example of a publicly-funded paper that you couldn't find on PubMed or another server, but a Nature or another closed journal offers for sale?
Click the first research article they show. It has various public funding sources:
"The collection of the ELC and genetic data for the American samples was supported by direct funding from the Intramural Research Program of the NIMH to the Clinical Brain Disorders Branch (D.R.W., PI, protocol 95-M-0150, NCT00001486, annual report number: ZIA MH002942-05), with supplemental analytic support from the Clinical and Translational Neuroscience Branch (K.F.B., PI). G.U. received partial support from P50MH094268."https://www.nature.com/articles/s41591-018-0021-y
I don't have the time to dig up an example, but it is a fairly common problem to run into. I went through the exercise of reviewing everything I could find related to a relatively specific business problem (marketing attribution) a few years ago and there were several papers that were locked behind a paywall without a free version available.
No. I usually assume that a paywalled article is just not (legitimately) available for free. But for the last paper I did want to read, I was able to find a free version.
> Are you claiming that most research is already freely available and yet everyone is, for some unspecified reason, still paying publishers for an extraneous service?
I'm not sure about the NSF, however: this is what is happening in some disciplines. ArXiv is the default method of sharing research in some fields, and also the default method of actually reading and reviewing those articles, yet people still submit their research to the traditional publishers, and people still either pay publishing or subscription charges to those publishers.
That's because the service those publishers provide is not the publishing or making accessible of research, but the academic credentials one needs to advance their academic career. That's the "unspecified reason", and it's an incredible waste of money.
I'm with you on all of what you wrote and that is my understanding too.
But you didn't explain, relative to how I interpreted the comment to which I initially replied, why, e.g. libraries, are paying so much money to access journals. If the default method for reading those articles is ArXiv, what's the point of paying a subscription to the other journals? My model of academia isn't that it's that dysfunctional. I'd expect enough internal pressure to re-allocate those funds if ('read') access was really unnecessary.
Right, that's an interesting observation as well. The number of journals is huge. What this means is that libraries usually do not buy individual subscriptions. Instead, what happens is that they negotiate so-called "Big Deals" with the big traditional publishers. In other words, they pay a large sum of money for a package of subscriptions to a large number of journals from a single publisher. Typically, these packages contain a few of the really big name journals, and a really long tail of journals of varying relevance. Publishers also own many journals which presumably mainly exist to boost those numbers in order to negotiate a higher price, and also has the effect of removing downwards price pressure by decreasing transparency in the market. Theoretically, the per-journal price is a lot lower in such a big deal than individual subscriptions would have been, but it's not quite clear how many of those subscriptions you really need.
Of course, "need" is already a vague term - preferably, all publicly funded research would be available in the first place, but that's not the world we currently live in.
Unfortunately, that's both the reason why adding a subscription journal instead of an Open Access one is bad, as well as why a boycott doesn't affect much change other than that resulting of its publicity. On the scale of those big deals, saying "we don't need that specific ML journal since interesting research hardly gets submitted there" doesn't drive the collective price down that much.
(In fact, I'd wager that those prices are mostly driven by how far librarians managed to stretch their budgets, rather than by the value provided by the specific included journals or the cost spent producing them.)
The parent very clearly was not saying that most research is already freely available. You're incorrectly claiming that. The parent said in most cases regarding US taxpayer funded research, right here:
> in most cases I know of (NSF and NIH funded research) US taxpayer funded research must be made freely available
Besides that, you'd be surprised that people and organizations spend money on things they don't actually need to? In my experience that happens constantly in both business and government.
I didn't incorrectly claim anything about the parent. I was asking them if they were claiming what I thought they were claiming. Did you miss the question mark at the end of my first sentence?
And it's pretty aggravating for you to quote the comment to which I directly replied. It's not even that long!
I was surprised by the implied claim, or what I thought was being implied, because I'd suspect "US taxpayer funded research" would include almost all research done in the U.S. and yet I'm pretty sure that in fact lots of people struggle to find freely available papers and articles and that lots of academic organizations pay lots of money for journal access. The parent seemed to imply that 'everything' really was freely available.
That papers will eventually (and mostly) be freely available is good and I am surprised that that seems to be the case. Thanks for sharing this info!
But an embargo or delay in publishing to the free access sites is probably sufficient for the existing closed-access journals to maintain their value. I can't imagine researchers, and thus universities, forgoing access to current research.
Yeah; I saw that in another comment. Still, this is very good news. Hopefully everything will be freely available someday soonish, and without embargoes or delays.
As long as they put the code on github or a blog or whatever then who really cares? Do people really get that much out of these academic style papers in this field? They are more like a press release.
Or are you saying this embargo also actively prevents them from sharing stuff beyond the document, like the code?
This is true, but with a couple huge caveats. NIH-funded papers published before April 7, 2008 are not applicable to the public access requirements. Research published using NSF grant proposals submitted before Jan 25, 2016 is also exempt. This encompasses the vast majority of all existing NIH and NSF-funded work, and includes many yet-to-be-published NSF papers. In addition, NSF-funded research does not need to be made public access until a full 12 months after initial journal publication.
Unfortunately, the issue is really complicated, yet almost everyone who learns more about it will come to the conclusion that the current publishing model is ridiculous. Simplification is an effective way to introduce the issue, and in this case doesn't even lead people to conclusions they wouldn't otherwise arrive at.
Ideally, of course, we'd never have to simplify, but that's life.
Mostly papers are thrown up on arxiv. I don't see why we should have to pay to host our own papers when we could just host our own papers. Seriously, three grand? How many months of AWS S3 does that pay for?
In my experience those high fees usually only apply when publishing open access. In our lab we publish using the free closed access method, and then a free version goes on PubMed (potentially after an embargo). So the research is available for free, the lab still gets the journal prestige, and it doesn't cost the lab anything. This is by no means a perfect system, but it's not as dire as a lot of these articles want us to think.
Only math/physics/CS papers are put on arxiv. These are a small fraction of the total academic and industrial research output. In my own field, earth science, there is an arxiv-clone (eartharxiv.org) that is a few months old. I bet the majority of earth scientists (and certainly 90% of geologists) have never heard of arxiv.
Furthermore, many people just put pre-prints on arxiv etc. and don't update after peer review. Peer review is painful at best and quite flawed but the median paper is actually better after it, and many papers rightly don't make it through. Having open-access curated journals makes a lot of this better.
Because as a researcher, you're not paying to have your papers hosted or even to have the feedback process facilitated, but you're paying so you can say your articles have been published in <journal x>, which helps you when applying for grants or tenure.
Most journals are OK with authors posting a paper on their own website, especially the 'pre-print' (pre-review, pre-layout). See: http://www.sherpa.ac.uk/romeo/index.php
(even Nature is OK for all the pre-prints, arxiv and your own website).
Former scientist here, too. People tend to forget what value a Nature article carries for the successful career in academia. The technical “hosting” of Nature and Arxiv might be the same, but the first will buy a lot of fame and potentially a good position at an research institute.
No one forgets that. Everyone just thinks we should solve the chicken-and-egg problem and move that prestige to a set of journals that don't gouge everyone for research that was mostly paid for by public money already and was peer-reviewed by public employees volunteering.
this has nothing to do with some mystical 'nature of humans', we're just stuck in a bad nash equilibrium.
Journals thrive on reputation and nobody can individually dissent because they lose out. This is the same mechanism that keeps overpriced tuition fees at ivy league colleges afloat.
The answer is to collectively get out by legislating the practice or offering a free public alternative that everybody can immediately jump too. Given that the efforts become more and more organised, (German wholesale boycott of Elsevier) this might be happening sooner than later.
Just offering a superior platform is not enough, as said in the last post, you also must create a way to collectively move everybody over and break the feedback loop of prestige journals.
The same happens with social networks. You could offer the freest, most privacy respecting network of all, as long as individuals get punished for leaving the old one individually it'll never take over a significant amount of market share, even if everyone on the old network agrees that the newer one is better.
I'd say arxiv isn't a superior platform. It is in that it is free, but it is the filtering effect of peer review that seems important. I'd say you don't need to actually filter, just some sort of measure of quality to make searching easy. I look for papers these days by hopping links on google scholar citations. This works well for finding good "older" papers. I want a mechanism for quickly finding good new papers. Then you have your superior platform.
Everybody would agree on this - in theory. But you don’t say how this should happen, and this is where it breaks down. The only thing that comes to mind is “magic”...
... which is exactly our problem. Everyone (bar the publishers) hates the status quo, but nobody knows how to break it. My personal hope is that sci-hub will gradually erode the profitability of scientific publishing down to nothing, but breaking the stranglehold of publishers will almost certainly be a long and painful process.
1. The governmemt could require you to publish all your research in an open way, if it ever funded anything to do with your research. No open access, no government money.
2. Universities could band together and do something about it.
This could happen both on the demand side, AND on the supply sude. For example, they could either collectively refuse to pay for journals, OR collectively refuse to hire professors who don't publish open access.
Want to get Tenure? Well you are forced to publish open access. Don't like it? Then go find a different University to work for. But that might be difficult if those OTHER universies collude against you with the same rules.
The problem (in this case) isn't "social status currency". It's that the social status currency system has real (financial) costs too!
The value of a prestige journal brand is that it is (supposedly) selective in which papers it publishes; not that it costs a lot of money for people to read those papers.
One thing I really wonder is wheather those costs are actually necessary for simple online publications:
- Peer review is basically "free", meaning it's done by volunteers without compensation.
- Hosting cost is magnitudes lower than publication/reading fees. A few dollars per paper should be enough by far.
- Organisation of peer reviewing, editing and so on imposes some costs, but a lot of it could be automated and the remaining cost per published paper should still be quite low.
So, would it be technically possible to create such journals? Of course, prestige might be a hurdle, but given some high-profile researchers would join as editors initially, that could work out.
I have heard the organization, editing, and publication cost is underestimated because it's really boring work and not as automatible as you might think.
But it's still not high enough to justify journal subscription prices.
I don't know if typical open access journal fees are lower margin or not.
The challenges are really not in the publishing infrastructure. Your assumption that some (where "some" really should be "pretty many" in absolute numbers) high-profile researchers would join does not seem likely, unfortunately. For some examples by some people who are spending a lot of effort trying to convince editors to do so: https://gitlab.com/publishing-reform/discussion/blob/master/...
I'm actually more concerned about the negative (your statement is mostly true), i.e. "if it's in a second-tier journal, it doesn't exist". Sure, some just look for an outlet for their crappy research, but the publication outlet shouldn't be used as a shortcut to evaluating quality.
Journal reviewing is oversold in my opinion. Just like Wikipedia became a better encyclopedia than traditional encyclopedias, I don't see why open-access approaches can't be at least as efficient at truth-seeking and consesus-building than traditional journals.
The whole idea behind journals is peer-review. That's very nearly a wikipedia-style open access system anyway. There's just not a great reason to keep chargind exhorbitant publishing fees when it's published online.
They often are at least as efficient and truth-seeking. Unfortunately, researchers do not pick journals on the basis of who efficient and truth-seeking they are, but on how much they can help their career. (I mean "unfortunately" here in the sense that the circumstances are unfortunate - I can't blame the researchers.)
A bit off topic but have research journals become a whipping boy for the anti-science crowd? The general public is being spoon fed either very poor quality click-bait "break through/cure cancer" articles or worse. Recently the peer-reviewed != reproducible issue was highlighted. I feel there is a common thread but it could my commonality bias of an HN reader. Is attacking science journals a new trend?
Supporting reproduction, complaining about public presentation of science, and disliking the rent-seeking behaviors of journals are all pro-science positions held more often by scientists than any other group.
May be these journals should move away from publishing and rather be in the business of curation from a source such as arxiv; Arxiv has a wealth of content, may be journals could sift through them and produce periodic curated lists of work that is the most significant.
"setting it to download academic journal articles systematically from JSTOR using a guest user account issued to him by MIT"
INFORMATION IS FREE:
"Federal prosecutors later charged him with two counts of wire fraud and eleven violations of the Computer Fraud and Abuse Act, carrying a cumulative maximum penalty of $1 million in fines, 35 years in prison, asset forfeiture, restitution, and supervised release."
It's really strange to me that in 2018, with open web and were we are publishing massive codebases on github that folks are still going through a paid/closed publisher to get published.
While many eyes may make all bugs shallow; getting those eyes isn't free.
That is, the primary value-add of the traditional paper publishing model is trust in the editorial staff and peer-reviewers. Anybody can throw up a repo on github - not everyone can get that repo published in Nature. Journals provide a curated list of exciting, valuable, and reliable research.
While we are on the topic, why are we supporting IEEE? They have so many valuable as well as historical papers behind their paywall. I'm IEEE member and want to find out what can I do about this.
IEEE to me is a different case. They do more than publish and IEEE, along with SPIE, although expensive still carries weight for me. I can assume the content is reasonably high quality and well reviewed.
In general, though, for-profit publishing will hopefully go away soon. As an engineer I need quality scientific publications without the distorted politics behind it. I could come up with any number of ways to improve on the current situation.
Nature used to be a prestigious journal. But its spinoffs, not so much. When they stuck to bio, they were good. Computer science, not so good. Nature Energy's hype pieces from flaky battery inventors have been mentioned on HN.
Journals are based on prestige; researchers give their work for free to a publisher where it is reviewed by other researchers (usually for free) and then published such that other researchers must pay to get the work. It does create a well respected system since publishers want only quality work but the arrangement has become more and more one sided favoring the publishers.
TL;DR This isn't that surprising. The cost of publishing is lower than ever and researchers are tired of getting charged for what they give away for free.
I have a thought... Force car owners, business or consumer, to buy driverless car insurance. No insurance, no autonomous vehicle functionality. Insurance companies would love a new market opportunity, and they'd lead the push to make driverless cars safe enough to insure.
Honestly, they should just flip the journal funding model on its head. Instead of readers paying for subscriptions, publishing institutions can pay for publication rights. In fact, they could probably take all the subscription money they get from institutions and change nothing but the name (something like "authorship fee", "membership fee" or "institutional publication service fee" or whatever). make the journals open access and not feel any change at all in revenue.
Besides, if you look at the money, fancy journals are responsible for a lot of funding that ends up in the coffers of researchers or the institutions that support them, so it really wouldn't be crazy. I read somewhere that a Science/Nature paper in some fields is thought to be worth >$1mm in funding.
I don't know for sure, but I'd guess that the vast majority of NPG's revenue comes from institutional subscriptions and that the vast majority of those institutions make use of NPG's review/editorial process.
I guess there's less of a notion of the accumulation of valuable property in the pay to publish regime.
But seriously, just call the subscriptions "membership fees." Revenue would probably dip 5% from non-publisher subscriptions and things would stabilize.
It would also neutralize the threat of open access journals gaining more ground.
What's the point of journals these days, especially in computing with reproducible results (i.e. authors publish a paper on arxiv, code & dataset on github)?
How are you going to change the leadership of a company you don't own? What are you planning we do? Get together and buy 51% of RELX? It's a $33 billion company.
Copernicus Publications publishes a host of journals, many on the behalf of non-profit scholarly and scientific societies [1]. For some or all of their journals, including those by the European Geosciences Union, peer review is open and there is a month-long public discussion forum for each manuscript.
I've got a paper in review with one of the journals right now, but several of my colleagues have expressed caution with open peer review. I think a lot of scientists are worried about making work public that hasn't been privately vetted during peer review, and maybe being caught with an embarrassing mistake.
If people can post a pre-print to Arxiv then why is it a big deal if it also gets published in Nature Machine Intelligence? Sure a better formatted and edited version may be behind a paywall but anyone can find the original arxiv paper and get the important points.
edit: To make it clear, I'm talking about submitting to BOTH Arxiv, where it's freely available, AND a paywalled journal like Nature.
edit 2: I get that we don't want to waste, often government, money on mere distribution of research, but I also think researchers need a way to distinguish their work from the thousands of arxiv papers that get posted every year. Whether that's through a Nature publication or something else is immaterial.
Academic journals don't result in any value creation for society. If "publishing in Nature" is no longer a credential for academics, then Nature can't charge universities big bucks... that money can be spent by institutions on other things (like funding GPU's to do more actual machine learning research)... so boycotting the journal is strictly a good thing.
This is absolutely wrong. Academic journals provide value through scientific aggregation, review, editorial curation, publication, archival curation, and promotion.
Yes, you can bypass all of this and put your non-reviewed pdfs on your personal website. No one will care, in part because no one will even know and if they do, they have to do the hard work of review in order to decide whether a paper is worth trusting. But how will anyone find out?
As an author, I've made a point never to publish in a paper not backed by a non-profit scientific society with a decent open-access policy, after doing so in the past. I also put all of my pdfs on my website including appropriate white papers that are working drafts, making it clear to others that they haven't been accepted through peer review.
As a reader, however, I keep up with the literature (aside from targeted Google Scholar searches) through Table of Contents emails from a few journals--mostly non-profit society journals but a few others including Nature. The editorial curation aspects which limit the number of accepted papers per issue make this possible. I'm not going to drink from the firehose of shit that is produced each month, especially before peer-review. Yeah, it sucks when your paper is accepted by reviewers but gets nixed by the editors because it's not cool enough, but... you know, get back to the bench and make it cooler if you want people to care. Welcome to the 21st century and attention scarcity.
You can change the name 'Nature' to whatever open-access journal you'd like, but the power behind it is that it's a sign that Important People have decided your paper is one of the 10 most impactful papers in your field that year. That's why people care.
Maybe I should have stated it as "paid academic journals don't create any marginal value for society vs. unpaid journals"
Nature is just a bunch of noted academics that come together and do peer-review (for free btw)... is that enough value creation to charge exorbitant amounts to universities and block content to the public that is funded by government grants?
If academics could build a peer-review board themselves (hint: now they can... thanks internet!) ... then everyone would have free access to science.
Universities in developing nations get heavily discounted if not free subscription and individual access fees to most journals. Library subscriptions are negotiated on a school-by-school basis and you can be sure that a university in sub-Saharan Africa isn't paying the same as Yale.
> Academic journals don't result in any value creation for society
Journals do provide some value. Editing, printing, and herding academics through the camera-ready process all require non-trivial effort.
I would have absolutely no problem with a handful of universities/non-profits/even for-profits devoting some staff to the curation/archival process. And also no problem with those organizations charging others modest fees to fund that work.
The problem with journals is that they capture a lot more value than they provide, that value is not distributed back to the people doing the actual work, and the prestige signal/branding provides a strong moat.
I REALLY think this is an important point. If we continue to behave like all of this stuff is trivial and obvious to replicate, then we'll be forever stuck with predatory journals like Nature.
Yeah, but think about all the trash they must reject if they let the questionable things trough.
I’m also more sceptic that the flagship journals actually have declined, at least in part I suspect the critsism is just complains in the spirit of those about ”the youth of today”.
Actually, some of the biggest problems with modern science are directly caused by journals.
One problem is that because they publish the most eye opening studies, they are inherently biasing the results they publish to extreme findings that are anomalous or don't represent reality. If 20 scientists are all conducting similar experiments, a typical experiment design would expect 1 of them to find a statistically significant result when there is none. But due to the way journals review and accept research, often the extreme result will be the one published. In fact, surveys have shown that many of those with the null result won't even bother submitting their findings because they likely won't be accepted. So published research is being biased towards false positive results, which is part of why reproducibility in science is in crisis these days.
So while there is certainly bad science being filtered out by journals today, there is actual important "boring" findings that are solid science that are being filtered out as well, and that is damaging science on the whole.
> One problem is that because they publish the most eye opening studies
Is this really the case? In my own field, it is quite common for us to publish only slight increments of existing knowledge, or we simply debunk an earlier paper. I have also done some work in scientific editing for scholars in other fields, and from what I’ve seen, it seems that they, too, publish similar minor work, even literature reviews where the author does not make any original claims but only summarizes what is out there, and these do get accepted and published.
I can definitely imagine that a few über-high-profile journals like Nature might want the flashiest claims and discoveries, but there is a whole world of publishing out there.
If you have two papers in NIPS every year then yes, your tenure at a big school or .com is assured and all is well. But that is less than the top 0.1% of active ml researchers, and also the most fashionable 0.1% as well. For the rest the impact factor is important in order to get another gig!
> Whether that's through a Nature publication or something else is immaterial.
But it's not really immaterial, is it? It's a ridiculously expensive way to sift through research, as there is hardly any downwards price pressure at all.
> In my own field of machine learning, itself an academic descendant of Gauss’s pioneering work, [...].
Yes, and I'm a descendand of Julius Caesar, Confucius and Charlemagne. I won't
tell you this when introducing myself though, because so is everybody else.
It wouldn't have occured to me that AI researchers have such inferiority complexes that they need to descend to such name-dropping, but I guess I was wrong.
I think he means that the field of ML has been built in part from Gauss's work on algebra and statistics, not that the author is related to Gauss himself.
The criticism is along the lines of "Yeah basically all fields are academic descendants of Gauss. Euler too." The article's aside is just odd, as is in the first place bringing up Gauss (linking to another Guardian article, maybe the real intent?) and an astronomer both paying another dude (some amount) what sounds like club member fees for letter redistribution.
Not a "legitimate" descendant, whatever that means after 60 generations.
However, he was an aristocrat, and a soldier, in a time when birth control was rarely practiced. I imagine he's got plenty of his DNA in the modern pool.
I work with a fair amount of people from academia, a lot of them seem to have the view that things that have not been peer reviewed are not worth reading.
But ML already has conferences and journals where work is peer reviewed and published in an open access manner.
So it's not that academics see journals/conferences as value-less, but that they are already doing all this work for free, why do it for free for Springer who will then charge everyone?
I think a lot non-academics (myself included) tend to hear the phrase "peer reviewed" and conflate it (rightly or wrongly) with "these results are replicable and therefore correct."
I don't know what the long term solution is but I have a feeling it won't involve companies that charge a lot of money for a service (publishing) that has been effectively free since the Internet has existed.
A lot of peer reviews in ML are currently completely open. For instance you can view the reviews for papers in previous and upcoming conferences here [0].
Also in terms of replicability ML is in a unique position where a lot of research is on standard datasets with models that don't take an excessive amount of compute power to train. This means that if the authors include a github link to their code you can fully replicate their results. So some "peer reviewed" ML papers can actually subscribe to your first description of peer reviewed.
I think a lot of ML research, despite being replicable is not necessary useful, because what we really want, approaches we can add in addition to all the other methods we have, that introduce meaningful improvements without unnecessary complexity and work across datasets, are quite rare. And that's assuming the evaluation was done well.
Not to say the research is worthless, just that straight replicability is not necessarily enough.
The greatest value of journals isn't that three random people (give or take) have reviewed and suggested edits to the papers before publication.
The greatest value of journals is when hundreds or thousands (or, hell, dozens) of other people, well informed and competent in the subject matter, read the papers and consider how it meshes with their own experience and understanding and build upon it, writing letters to the journal pointing out problems and articles which follow up on the original work.
That is how an idea is validated. An article published in a journal that no one reads isn't much different than an unpublished article, and if work is spread, responded to, and built upon outside of journals, it isn't much different from a journal-published article.
That's unfortunate because that is how consensus and dogmatic group-think emerge. People need to think for themselves.
I love learning but stayed away from college because I couldn't stand how fundamentally at odds traditional education routes are to creativity and free thinking.
Personally, I find Twitter and arxiv-sanity good sources for keeping up with ML. They have their own mechanism of surfacing popular content and filtering based on preferences.
That role (in machine learning) is already filled by conference proceedings (plus ArXiV), so this journal is an opportunistic attempt to monetize a niche which doesn't need filling.
In some fields, like machine learning, many researchers put their work onto arxiv. There is like 50 papers a day now. https://arxiv.org/list/stat.ML/recent
People don't find out about research by reading journals. They go to lectures and read mailing lists for that. Journal articles are typically a year too late.
The value-add from journals is that they (at least theoretically) guarantee another person has read the paper carefully and believes it is correct.
With biology at least a lot of your reputation relies on your reputation, and a lot of that comes from publications and the impact factor of the journals you publish in, Nature being the one to aim for. Not sure if its as relevant to AI.
Computer science doesn't need journals because (1) It's generally easy enough to reproduce any work you care about, so you don't need to take anyone's word for its correctness, and (2) researches can make a fortune in industry doing their (profoitable) work or something related, so the don't need to suck up prestige factories in order to get funding.
I was thinking as well that its less abstract. Biology research takes a lot time and of work to filer through to practical applications. AI research is generally doing something practical straight away.
The old model may be obsolete, or at least need tuning. What about offering a "vetting grade"? Anybody can submit (assuming minimum standards) and anyone can read, but you'd have to pay to have your research vetted, perhaps with different levels of vetting. Don't make it all-or-nothing.
Would like to see Demis Hassabis and David Silver join the effort.
If Deepmind, OpenAI and Berkeley (well represented here) all boycott, that's a huge chunk of key AI researchers.