The recent announcement to reject review articles and position papers already smelled like a shift towards a more "opinionated" stance, and this move smells worse.
The vacuum that arXiv originally filled was one of a glorified PDF hosting service with just enough of a reputation to allow some preprints to be cited in a formally published paper, and with just enough moderation to not devolve into spam and chaos. It has also been instrumental in pushing publishers towards open access (i.e., to finally give up).
Unfortunately, over the years, arXiv has become something like a "venue" in its own right, particularly in ML, with some decently cited papers never formally published and "preprints" being cited left and right. Consider the impression you get when seeing a reference to an arXiv preprint vs. a link to an author's institutional website.
In my view, arXiv fulfills its function better the less power it has as an institution, and I thus have exactly zero trust that the split from Cornell is driven by that function. We've seen the kind of appeasement prose from their statement and FAQ [1] countless times before, and it's now time for the usual routine of snapshotting the site to watch the inevitable amendments to the mission statement.
"What positive changes should users expect to see?" - I guess the negative ones we'll have to see for ourselves.
I came here to say something similar. As someone who works in a field that applies machine learning but is not purely focused on it, I interact with people who think that arXiv is the only relevant platform and that they don't need to submit their work to any journal, as well as people who still think that preprints don't count at all and that data isn't published until it's printed in an academic journal. It can feel like a clash of worlds.
I think both sides could learn from the other. In the case of ML, I understand the desire to move fast and that average time to publication of 250-300 days in some of the top-tier journals can feel like an unnecessary burden. But having been on both sides of peer review, there is value to the system and it has made for better work.
Not doing any of it follows the same spirit as not benchmarking your approach against more than maybe one alternative and that already as an after-thought. Or benchmaxxing but not exploring the actual real-world consequences, time and cost trade offs, etc.
Now, is academic publishing perfect? Of course not, very very far from it. It desperately needs to be reformed to keep it economically accessible, time efficient for both authors, editors and peer reviewers and to prevent the "hot topic of the day" from dominating journals and making sure that peer review aligns with the needs of the community and actually improves the quality of the work, rather than having "malicious peer review" to get some citations or pet peeves in.
Given the power that the ML field holds and the interesting experiments with open review, I would wish for the field to engage more with the scientific system at large and perhaps try to drive reforms and improve it, rather than completely abandoning it and treating a PDF hosting service as a journal (ofc, preprints would still be desirable and are important, but they can not carry the entire field alone).
Simply anticipating basic push backs from reviewers makes sure that you do a somewhat thorough job. Not 100% thorough and the reviews are sometimes frivolous and lazy and stupid. But just knowing that what you put out there has to pass the admittedly noisily gatekept gate of peer review overall improves papers in my estimation. There is also a negative side because people try to hide limitations and honest assessments and cherry pick and curate their tables more in anticipation of knee jerk reviewers but overall I think without any peer review, author culture would become much more lax and bombastic and generally trend toward engagement bait and social media attention optimized stuff.
The current balance where people wrote a paper with reviers in mind, upload it to Arxiv before the review concludes and keep it on Arxiv even if rejected is a nice balance. People get to form their own opinion on it but there is also enough self-imposed quality control on it just due to wanting it to pass peer review, that even if it doesn't pass peer review, it is still better than if people write it in a way that doesn't care or anticipate peer review. And this works because people are somewhat incentivized to get peer reviewed official publications too. But being rejected is not the end of the world either because people can already read it and build on it based on Arxiv.
> Unfortunately, over the years, arXiv has become something like a "venue" in its own right, particularly in ML, with some decently cited papers never formally published and "preprints" being cited left and right. Consider the impression you get when seeing a reference to an arXiv preprint vs. a link to an author's institutional website.
This just isn't true. arXiv is not a venue. There's no place that gives you credit for arXiv papers. No one cares if you cite an arXiv paper or some random website. The vast vast majority of papers that have any kind of attention or citations are published in another venue.
My observation is that research, especially in AI has left universities, which are now focusing their research to a lesser degree on STEM. It appears research is now done by companies like Meta, OpenAI, Anthropic, Tencent, Alibaba, among many others.
Universities (outside a few) just have much weaker PR machines so you never hear what they do. Also their work is not user facing products so regular people, even tech power users won't see them.
> raised concerns about the proposed $300,000 salary for arXiv’s new CEO, saying it seemed high
Is a mid-to-high engineering salary outlandish for a CEO of what is likely to be a fairly major non-profit? Even non-profits have to be somewhat competitive when it comes to salary, and the ideal candidate is likely someone who would be balancing this against a tenured position at a major university
The reason is because arxiv is growing significantly leading to 297,000 deficit in operating costs for 2025 alone. Corenell has helped with donation a long with other organizations that pay membership fees.
As a result, donors + leaders of arxiv think it's best to spin off to increase funding.
The "essentially static hosting" isn't the cost centre (although with 5 million MAU, it's nothing to sneeze at). The real costs are on the input side - they have an ingestion pipeline that ensures standardised paper formatting and so on, plus at least some degree of human review.
No, I mean that the pipeline requires software engineers to build/maintain, and salaries are (as in basically every tech organisation) the dominant cost
The PDF formatting is all but standardised. They ingest LaTeX sources, which is formatted according to the authors' whims (most likely, according to whatever journal or conference they just submitted the manuscript to).
I'll concede that the (relatively novel) HTML formatter gives paper a more uniform appearance. They also integrate a bunch of external services for e.g., citation metrics and cross-references. Still hard to justify such a high cost to operate, but eh.
Also, the "human review" is a simple moderation process [1]. It usually does not dig into the submission's scientific merits.
This sounds terrible. Of course there's a huge risk of it becoming made for-profit. It almost makes you wonder if the academic publishers are behind this push somehow.
Could they not have made it into some legal structure that puts universities at the top? Say, with a bunch of universities owning shares that comprise the entirety of the ownership of arXiv, but that would allow arXiv to independently raise funds?
> Of course there's a huge risk of it becoming made for-profit.
The article says that "it will become an independent nonprofit corporation", and as OpenAI's failed attempt showed, converting a non-profit to a for-profit organization is either really hard or impossible.
> Could they not have made it into some legal structure that puts universities at the top?
As a corporation (even a non-profit one), it will have a board of directors. I have no idea what their charter will look like, but I would be surprised if at least one seat wasn't reserved for a university representative, and more than that seems quite likely as well.
OpenAI didn't get everything that they wanted, but I very much disagree with calling it a "failed attempt". The non-profit went from owning the entirety of OpenAI to having ~25% stake.
I fear their Mozilla-ification and Wikipedia-ification. Scope creep, various outreach feel-good programs, ballooning costs, lost focus etc. And other types of enshittification.
Any change to the basic premise will be a negative step.
They should just be boring quiet unopininionated neutral background infrastructure.
I've often thought that similar trust systems would work well in social media, web search, etc., but I've never seen it implemented in a meaningful way. I wonder what I'm missing.
Now the question is, will arxiv wage a decade long bloody war with Cornell, using heavy infantry (PhD students), archers (reviewers) and field artillery (AI slop papers), or will the independence be mostly peaceful? Only time can tell.
And they hired a LinkedIn business idiot to run the new organization - so the aim is for an infinite growth tech startup in terms of governance, despite the technical legal status of non-profit. It shows in the language they use in the announcement, too ("improved financial viability in the long run")
OpenAI shows exactly how well that works and what that kind of governance does to a company and to its support of science and the commons.
"Recently arXiv’s growth has accelerated. Since 2022, it has expanded its staff to 27, in large part to deal with a 50% increase in submitted manuscripts."
I am wary of that. IMO the business model is damaged therein. You can say in 2022 we had 27; bankrupt in 2030.
With 300K for the CEO, its enshittification will commence imminently. It will now serve to maximize revenue. Just wait and watch while they issue a premium membership, payment requirements for authors, and other revenue generators to please their investors.
they'll just turn into a shitty journal at this point, they just need to introduce peer review and they can start competing with the real journals on price point.
It is just a preprint repository. It is pretty open (the stories where a preprint was rejected or delayed unreasonably are extremely rare). It offers the basic services for a math/compsci/physics themed preprint repository.
I don't see much of a monopoly, nor any "moat" apart from it being recognised. You can already post preprints on a personal website or on github, and there are "alternatives" such as researchgate that can also host preprints, or zenodo. There are also some lesser known alternatives even. I do not see anything special in hosting preprints online apart from the convenience of being able to have a centralised place to place them and search for them (which you call "monopoly"). If anything, the recognisability and centrality of arxiv helped a lot the old, darker days to establish open access to papers. There was a time when many journals would not let you publish a preprint, or have all kinds of weird rules when you can and when you can't. Probably still to some degree.
It can host large datasets as well, yes. It is hosted by CERN, so it is not specifically IT in any way. It also allows you to restrict access to the files of your submission. It has no requirements to submit your LaTeX sources, any PDF will be fine. There are also no restrictions on who can publish. You'll get a DOI, of course.
Everything published on arXiv could also be published on Zenodo, but not the other way around.
It does do a fair amount of filtering of submissions, and it's a long term archive (e.g. for the next 100+ years). I suspect both (but with the former dominating) are the issue.
Well, technically, it can also compile your tex file if you upload the tex file instead of the pdf directly, which helps a lot in standardizing the stylistic structure between preprints. Most other repositories are wild west and inconsistent. I really appreciate the similarity in style applied to most preprints there. Moreover, this means you can also download not just the pdf, but the source tex file to, which can be very useful.
The similarity in style comes from conference and journal templates, not from Arxiv. You can style your paper with latex in any style, Arxiv doesn't care. On Arxiv you mostly see preprints that people submit to conferences and journals and they enforce the style.
Very unrelated to the article, but I think 'arXiv' as a brand is bad, and really detrimental to what the institution aims to accomplish.
That is, it's not readily parseable, it really gives an insider term vibe - like this isn't for you if you don't already know what it means or how you should read or say it. It sort of reminds me of the overuse of latin and latinate terms generally in the old professions and, well, the academy.
Just always struck me as being somewhat at odds with the goal.
I wonder what makes you feel that. I've been publishing preprints close to a decade on arxiv now and never had any particular feelings about it.
To me it's just a way to get out your work fast, so that there is already a trace of it on the Internets - nothing more and nothing less.
> That is, it's not readily parseable, it really gives an insider term vibe...
Isn't that normal with highly specialized research fields? I agree many papers could benefit from clearer wording, but working in a niche means you sometimes don't reach a broader audience
It's an opinion, and you feeling no particular way about it is equally valid.
But I did justify and maybe to reword slightly, surely if one of the main drivers is opening up research, the brand name should be something that's less obscure and more accessible / understandable as to what it is on first sight?
Maybe arXiv evoking the word 'archive' with an ancient Greek twist does that for some, but it's clearly a bit cryptic for many, and if the point is to open up probably the brand should just be something much plainer.
I've never even connected the 'X' to the Greek letter chi. I just kinda accepted it as one of many groovy web 2.0 misspellings in search of a domain and trademark.
everyone has a first time they see a thing and don't yet know what it is.
Using a brand as a filter where you have to already know what it means to get it is exactly the opposite of what it's supposed to achieve.
Consider the most exclusive (successful) brands that exist. Even there, where exclusivity is a brand goal, none of them have this property of being obscure on first contact.
You usually get introduced to it by your academic supervisor or collaborators as a masters or PhD student. If you're a solo researcher who has made a significant contribution on the frontier of science, I'm sure you'll be able to understand how Arxiv works as well. Because I assume you have had some conversations with other experts in the field. If you're a full on autodidact with no contact to any other researchers in the field, well, maybe it's better if you chat with some other people in that field.
Its reasonable to have a tradeoff here to avoid cranks and now AI psychosis slop. You can still post on research gate and academia.edu or you own github page or webhosting.
It's a classic story of someone having to pick a name quickly, which then gets established long before anyone who cares about branding is aware of its existence.
The original service didn't even have a name, only a description, and it was amusingly hosted at xxx.lanl.gov. But LANL wasn't really interested in it, and the founder eventually left for Cornell. At that point, the service needed a domain name, but archive.org was already taken.
And besides, the name has Ancient Greek influences. A similar Latinate term might be something like "archive".
The vacuum that arXiv originally filled was one of a glorified PDF hosting service with just enough of a reputation to allow some preprints to be cited in a formally published paper, and with just enough moderation to not devolve into spam and chaos. It has also been instrumental in pushing publishers towards open access (i.e., to finally give up).
Unfortunately, over the years, arXiv has become something like a "venue" in its own right, particularly in ML, with some decently cited papers never formally published and "preprints" being cited left and right. Consider the impression you get when seeing a reference to an arXiv preprint vs. a link to an author's institutional website.
In my view, arXiv fulfills its function better the less power it has as an institution, and I thus have exactly zero trust that the split from Cornell is driven by that function. We've seen the kind of appeasement prose from their statement and FAQ [1] countless times before, and it's now time for the usual routine of snapshotting the site to watch the inevitable amendments to the mission statement.
"What positive changes should users expect to see?" - I guess the negative ones we'll have to see for ourselves.
[1] https://tech.cornell.edu/arxiv/
I think both sides could learn from the other. In the case of ML, I understand the desire to move fast and that average time to publication of 250-300 days in some of the top-tier journals can feel like an unnecessary burden. But having been on both sides of peer review, there is value to the system and it has made for better work.
Not doing any of it follows the same spirit as not benchmarking your approach against more than maybe one alternative and that already as an after-thought. Or benchmaxxing but not exploring the actual real-world consequences, time and cost trade offs, etc.
Now, is academic publishing perfect? Of course not, very very far from it. It desperately needs to be reformed to keep it economically accessible, time efficient for both authors, editors and peer reviewers and to prevent the "hot topic of the day" from dominating journals and making sure that peer review aligns with the needs of the community and actually improves the quality of the work, rather than having "malicious peer review" to get some citations or pet peeves in.
Given the power that the ML field holds and the interesting experiments with open review, I would wish for the field to engage more with the scientific system at large and perhaps try to drive reforms and improve it, rather than completely abandoning it and treating a PDF hosting service as a journal (ofc, preprints would still be desirable and are important, but they can not carry the entire field alone).
The current balance where people wrote a paper with reviers in mind, upload it to Arxiv before the review concludes and keep it on Arxiv even if rejected is a nice balance. People get to form their own opinion on it but there is also enough self-imposed quality control on it just due to wanting it to pass peer review, that even if it doesn't pass peer review, it is still better than if people write it in a way that doesn't care or anticipate peer review. And this works because people are somewhat incentivized to get peer reviewed official publications too. But being rejected is not the end of the world either because people can already read it and build on it based on Arxiv.
It is an interesting instance of the rule of least power, https://en.wikipedia.org/wiki/Rule_of_least_power.
This just isn't true. arXiv is not a venue. There's no place that gives you credit for arXiv papers. No one cares if you cite an arXiv paper or some random website. The vast vast majority of papers that have any kind of attention or citations are published in another venue.
Is a mid-to-high engineering salary outlandish for a CEO of what is likely to be a fairly major non-profit? Even non-profits have to be somewhat competitive when it comes to salary, and the ideal candidate is likely someone who would be balancing this against a tenured position at a major university
The reason is because arxiv is growing significantly leading to 297,000 deficit in operating costs for 2025 alone. Corenell has helped with donation a long with other organizations that pay membership fees.
As a result, donors + leaders of arxiv think it's best to spin off to increase funding.
I've contracted into some consultancy teams which you could uncharitably describe as "15 people and $4mn/yr to create one PDF per month".
Most people I talk to hate that pipeline and spend a lot of debug hours on it when Arxiv can't compile what overleaf and your local latex install can.
Also, the "human review" is a simple moderation process [1]. It usually does not dig into the submission's scientific merits.
[1] https://info.arxiv.org/help/moderation/index.html
Could they not have made it into some legal structure that puts universities at the top? Say, with a bunch of universities owning shares that comprise the entirety of the ownership of arXiv, but that would allow arXiv to independently raise funds?
The article says that "it will become an independent nonprofit corporation", and as OpenAI's failed attempt showed, converting a non-profit to a for-profit organization is either really hard or impossible.
> Could they not have made it into some legal structure that puts universities at the top?
As a corporation (even a non-profit one), it will have a board of directors. I have no idea what their charter will look like, but I would be surprised if at least one seat wasn't reserved for a university representative, and more than that seems quite likely as well.
Any change to the basic premise will be a negative step.
They should just be boring quiet unopininionated neutral background infrastructure.
Exactly. It should be a utility. Not quite dumb pipe, but not too far either.
You need your favourite academic gatekeeper (= thesis advisor) to vouch for you in order to be allowed to upload.
Then AI slop gets flagged and the shame spreads through the graph. And flaggings need to have evidence attached that can again be flagged.
> arXiv requires that users be endorsed before submitting their first paper to arXiv or a new category.
[1] https://info.arxiv.org/help/endorsement.html
It's probably not perfect but in practice, it seems to have been enough to get rid of the worst crackpotty spam.
OpenAI shows exactly how well that works and what that kind of governance does to a company and to its support of science and the commons.
TL;DR, it's fucked.
I am wary of that. IMO the business model is damaged therein. You can say in 2022 we had 27; bankrupt in 2030.
another will need to rise to take its place.
To this end, they added an endorsement requirement this year: https://blog.arxiv.org/2026/01/21/attention-authors-updated-...
People keep falling into the same trap. They love monopolies, then are shocked when those monopolies jerk them around.
I don't see much of a monopoly, nor any "moat" apart from it being recognised. You can already post preprints on a personal website or on github, and there are "alternatives" such as researchgate that can also host preprints, or zenodo. There are also some lesser known alternatives even. I do not see anything special in hosting preprints online apart from the convenience of being able to have a centralised place to place them and search for them (which you call "monopoly"). If anything, the recognisability and centrality of arxiv helped a lot the old, darker days to establish open access to papers. There was a time when many journals would not let you publish a preprint, or have all kinds of weird rules when you can and when you can't. Probably still to some degree.
Everything published on arXiv could also be published on Zenodo, but not the other way around.
That is, it's not readily parseable, it really gives an insider term vibe - like this isn't for you if you don't already know what it means or how you should read or say it. It sort of reminds me of the overuse of latin and latinate terms generally in the old professions and, well, the academy.
Just always struck me as being somewhat at odds with the goal.
To me it's just a way to get out your work fast, so that there is already a trace of it on the Internets - nothing more and nothing less.
> That is, it's not readily parseable, it really gives an insider term vibe...
Isn't that normal with highly specialized research fields? I agree many papers could benefit from clearer wording, but working in a niche means you sometimes don't reach a broader audience
But I did justify and maybe to reword slightly, surely if one of the main drivers is opening up research, the brand name should be something that's less obscure and more accessible / understandable as to what it is on first sight?
Maybe arXiv evoking the word 'archive' with an ancient Greek twist does that for some, but it's clearly a bit cryptic for many, and if the point is to open up probably the brand should just be something much plainer.
Using a brand as a filter where you have to already know what it means to get it is exactly the opposite of what it's supposed to achieve.
Consider the most exclusive (successful) brands that exist. Even there, where exclusivity is a brand goal, none of them have this property of being obscure on first contact.
Its reasonable to have a tradeoff here to avoid cranks and now AI psychosis slop. You can still post on research gate and academia.edu or you own github page or webhosting.
The original service didn't even have a name, only a description, and it was amusingly hosted at xxx.lanl.gov. But LANL wasn't really interested in it, and the founder eventually left for Cornell. At that point, the service needed a domain name, but archive.org was already taken.
And besides, the name has Ancient Greek influences. A similar Latinate term might be something like "archive".
Isn't that actually kindof a good brand signal for a repo of very specialized papers? "Fun with learning" in comic sans wouldn't help credibility.