A Brighter Future for Academic Publishing

Sound clip: You’re listening to a Times Higher Education podcast.

Miranda Prynne: Hello, welcome to the Campus podcast with me, Miranda Prynne, editor of Campus. We are delving into the complex world of academic publishing this week. This is a system fraught with controversy that is widely criticised and yet one that even now holds the key to academic success or failure. Despite calls for change, most in universities acknowledge that a culture of publish or perish still dominates in many corners of academia. And while the basic premise of researchers getting their work out there into the world to be read, analysed and developed further by other scholars is a good one, with the huge pressure to publish comes inevitable problems. The big commercial publishers are regularly condemned as exploitative. The integrity of more and more papers is being called into question with retractions soaring. Peer review is accused of being corrupted. And even open access for all its lofty ambitions is awash with complications. And yet, and yet, scholars want to publish and we all benefit when new knowledge is shared. So, I spoke to two academics who are carving different paths through the quagmire, seeking publishing solutions that will encourage and underpin quality research practices. First, I spoke to Paul Ayris, who is pro-vice-provost for library services at University College London, and who, 10 years ago, founded UCL Press as the UK’s first fully open access university press, where he remains chief executive. I learned how this now flourishing press was conceived and where university presses fit into the wider ecosystem. We also talked about what new publishing and peer review models Paul and his team are exploring to support better and more accessible dissemination of knowledge. I then spoke to Philipp Koellinger, a professor of social science genetics at Vrije Universiteit in Amsterdam, who is also the co-founder and CEO of a tech startup, DeSci Labs, which aims to help researchers share and access their work without paywalls or publication charges through an open peer-to-peer network. Now, I asked what compelled Philipp to launch this platform, and we talked through some of the key challenges associated with predominant publishing models and how DeSci seeks to plug a gap in the academic publishing landscape by enabling researchers to share work at each stage of its development, which brings multiple benefits you will hear more about shortly. But first, let’s turn to my conversation with Paul.

Miranda Prynne: Hi, Paul. How are you doing on this hazy January day in London?

Paul Ayris: I’m doing really well, thanks very much. Busy term with lots going on, so plenty to keep me occupied.

Miranda Prynne: Yes, I can imagine that seems to be the general status quo most of the time across higher education. Well, thank you so much for finding the time to talk to me today. Now, you founded UCL Press back in 2015 as the UK’s first fully open access university press. What led to this?

Paul Ayris: When I was starting to think about the future role of the library in the 21st century, I realised that libraries not only store knowledge and catalogue knowledge, they can actually create knowledge. And the best way to create knowledge and to share it is to have your own university press. UCL was the only major research-led university in the UK at that time not to have a university press, and that didn’t seem right to me. And I was also aware that the current long-form monograph format struggles in a commercial publishing environment. It’s very expensive. Sales are actually quite low, very low, in fact, on a global scale. And it doesn’t look like a format that will live in the future. So, for all those reasons, I thought open access press run as part of the library or founded within the library, putting UCL on the stage as far as dissemination of our research outputs were concerned. So, it was a mixture of all those feelings.

Miranda Prynne: Yes, I can see that there are multiple benefits there. I know it was 10 years ago now, which I still find the fact that 2015 is a decade ago slightly terrifying. But could you drag your memory back there and tell us a bit about how you actually went about setting it up and perhaps extract some advice for others who might be considering setting up an institutional press or independent publishing platform?

Paul Ayris: So, as I said, we had no commercial press using the UCL imprint. So, I was starting with a tabula rasa, completely, you know, it was a completely clear slate on the model that we could adopt. I went to talk to our vice-provost research, who was then Professor David Price, and to deans of faculty, particularly arts, humanities, and social sciences, because I wanted to start our work with a long-form book monograph format. And those are the academic areas that produce the most long-form outputs. So, I took soundings from all these colleagues, and then I started trying to devise a model that would work, that would meet all their requirements for scholarly impact, for very strict, clear review processes, and of course, the business model that underpins the work. I think the most important thing that I would say to anyone considering to start setting up an institutional press is that you need to be sure, A, that you have academic support for what it is you’re trying to do, that it fits into the university’s mission, what it sees itself doing in the research space, in the innovation space, and in society. And second, that you have the right investment and the ongoing budget to fund the activity. Otherwise, it’s going to fold after a year, 18 months, and nobody wants that. So you’ve got to plan for the long term and make sure that the academic support and the funding to fund the infrastructure are in place.

Miranda Prynne: That makes sense. And you say your initial focus was on monographs and therefore the disciplines that were more focused on that. But since then, UCL Press now publishes journals, books across multiple disciplines. And that very much ties in with what you say, really. I mean, that’s a representation of UCL as a university more broadly and its interests.

Paul Ayris: Absolutely, yes. So we have three publishing platforms at the moment. The oldest and best established, simply through age and years of operation, is the monograph platform. We also have started a very successful e-textbook platform and are getting lots of interest from academic colleagues about publishing their own teaching material in open access digital form as electronic educational materials and outputs. And we’re trying to develop the idea of how we can scale this up so that that approach to educational outputs is present in every course which UCL offers for the student body. So that’s the second platform. And then the third platform, as you say, is journals. We’re bringing in a number of journals, commercial journals, well, academically led commercial journals who want to move to an open access platform. We have quite a healthy waiting list of journal colleagues who want to come in under the UCL Press umbrella. And we’re also experimenting with a new journal format, the megajournal, the UCL Open: Environment, is our first megajournal where it’s cross-disciplinary and interdisciplinary. So every possible aspect of the environment, which in the commercial publishing world would appear in niche journals in a number of areas across many different publishers. All this material in the megajournal is brought together under one between two covers, as it were. So, anything to do with the environment, finance, health, global warming, political issues, all these articles can be brought together in one place in the megajournal. And that’s the model that we’re testing at the moment.

Miranda Prynne: You’ve already mentioned that you identified that monographs were likely to struggle within a commercial publishing space. Now, in the last few years, I think it’s fair to say that academic publishing as a commercial entity has been in a permacrisis with editorial walkouts from journals becoming increasingly common. We’ve seen more papers than ever being retracted. Big publishers are repeatedly accused of extractive business models. Are university presses part of the solution to this?

Paul Ayris: Well, you’ve hit on some really important questions. So, of the three platforms that I’ve mentioned in relation to UCL Press, it’s the journal space where the commercial offering is most, I think, soundly established. Clearly, there are pressures in the journal market. The inability of many countries, particularly in the global south, to be able to afford to buy these titles because they don’t have the budget to do so, is a really serious issue when it comes to ensuring that research is equitable and anyone anywhere in the world who is a researcher in a particular field can make a contribution. If you can’t get access to the literature, then that’s going to be really difficult. The open access model that this university press is developing, there is a space for us to put a new concept into the landscape and see how that takes off with regard both to potential authors and potential users. As I said earlier, we have a number of established journals, mainly society journals or learned society journals, who want to come into the UCL Press fold and use the UCL Press type model. And that’s very encouraging. So we’re looking at expanding our list of journal titles on our journal platform as we speak. And the megajournal is our major attempt as a completely new model for how research outputs are disseminated, how they’re publicised, and how they can be made available to a global community in an equitable way so that everybody has the chance to read the research that’s being published. So I think universities have a really important role to play through their presses in changing the publishing landscape.

Miranda Prynne: Of course, and not only does it make it equitable in terms of those who wish to access this research, but it’s going to then also accelerate the generation of knowledge as more people are able to learn from the research and potentially then develop it further.

Paul Ayris: I certainly think there is an accelerator factor in material being available in open access because of the visibility that that research has. Let me give you just one example. Our most accessed and downloaded monograph is by Professor Danny Miller, who’s a professor here in anthropology. And he and his colleagues on the back of EU funding from EU grant published a book on social media called How the World Changed Social Media. Now, that one book has been downloaded and accessed over half a million times in its lifetime as part of the UCL Press stable. That’s half a million accesses. What are the traditional figures for sales of hard copy over the bookshop counter? About 300 copies. The sector would say, well, if you sell 300 copies of a book across the bookshop counter, that’s a good outcome. I think that’s an appalling outcome, actually. And when you look at what open access can do to accelerate the dissemination of knowledge, to make it available to a wider population, there are all kinds of opportunities that open access delivers that the current model, the conventional commercial model, isn’t able to address.

Miranda Prynne: Yes, 300 copies to half a million is, um, I mean, they’re not, they’re not really comparable figures, are they? They’re worlds apart.

Paul Ayris: And we say the interesting thing is, Miranda, we still sell hard copies. So the default form for the monograph is to access it free of charge as an open access output digitally from a number of platforms around the world where we store our outputs. But you are able if you’re a reader and you want a paper copy, you can order a print on demand paper copy and we still sell paper copies in numbers very similar to that 300 number that I mentioned in relation to conventional commercial sales. But the added value is that the digital, which has tens of thousands of downloads and accesses, is available as well in the open access form and it’s free at point of use.

Miranda Prynne: So you’ve added an enormous amount of value simply by adopting the open access model. How does your open access business model work? Because there seem to be different approaches to this. Again, I wonder if there are some lessons here for others listening.

Paul Ayris: Well, I don’t think there’s one model that fits every institution. So, my own view is that you need to work within the context that you’re based and come up or use one of the business models that are available in the open access community that best meets your needs. Here in UCL, we get funding from the UCL research budget to fund the development of the press. It’s a tiny, tiny percentage. I think less than 1% of UCL’s total research budget is spent on developing and maintaining the press on an ongoing basis. That means that UCL authors can publish for free, provided they pass the normal peer review requirements for their outputs. We charge currently an APC or a BPC in this case for monographs for external authors if they’re not funded by a research funder. But we have a waiver programme as well, where we’re able through the income that we derive from the sale of print copies, we’re able to waive the book processing charges for a number of authors who come from developing countries or come from universities who couldn’t possibly afford the book processing charge. Our aim is to phase out the book processing charge in the longer term and have a fully diamond open access approach to funding the monograph output. At the moment it’s very close to being diamond open access, but not entirely free of fees that we charge external authors while we develop our income streams in other areas to enable us to move to a diamond fully diamond model.

Miranda Prynne: You mentioned peer review just in reference to UCL authors needing to have gone through that process before their pieces are published. Now I’m keen to talk about this because this is another area of concern that is widely expressed across higher education. There are worries that this system is in some areas being corrupted and is not always a reliable quality control method. So I wondered how UCL Press goes about trying to ensure quality and integrity of the published research that you put out.

Paul Ayris: That’s a good question. We have different models depending on whether we’re talking about monographs, research monographs or journals. So let me start with research monographs. We have a two-level approach to peer review for research monographs. When the author submits a proposal to publish a book, they’re invited to name potential peer reviewers from around the world who we could approach to ask for a review of the manuscript or the proposal if it’s still at proposal stage. We also have an academic editorial board, which I chair, made up of UCL academics who are also able to name potential referees for any manuscript or proposal. We get two written referees reports back from our chosen referees, and then their reviews are reviewed and assessed by the editorial board. So there’s two layers of assessment. And if we agree with the peer review reports, then we would issue a contract to publish. If there are any questions, we would approach the authors again and ask for clarification or more work in certain areas. And we found that double lock, as it were, the external peer reviewers and our internal editorial board be quite a sound one in terms of identifying the best proposals that come into us because obviously we get more proposals than we are able to publish in any one year. For journals, we well let’s look at the megajournal because we have set up a completely new system of open peer review for UCL Open: Environment because we want to test that as a new model for journal publication. So when an article comes in to UCL Open: Environment, the editor of the journal will have a quick look at the content and if he or she feels that it is suitable, it will go immediately to a preprint server, not be published, but sit on a preprint server as a green open access preprint and it will always stay there as a preprint. And if the author sends corrections or additions or wants to delete particular sentences, the second version will stay there as well on the preprint server. It will then go to peer review and we would look for two positive peer review reports and we publish the names of the reviewers and their affiliations so that who wrote the review, the reviews are published as well. So, who wrote the review is known to the audience. And then, when there are two positive, at least two positive peer reviews, it will be published in the published version of the journal. We found that using open peer review, it’s novel, but it is attracting interest, certainly from our authors who find it a very transparent way in which their work is being assessed and evaluated when it comes to publication. So, we’re going to carry on with that model and continue to monitor author reaction and reader reaction to that, those quality control mechanisms.

Miranda Prynne: Interesting. So, when you talk about open peer review, that’s in reference to the reviewers’ names and their institution. Which, of course, means everyone is held to account. The process you’ve described there sounds fairly rigorous.

Paul Ayris: I was really clear, and my academic colleagues who form the editorial board here for our monographs were absolutely clear right from the start that we would not sacrifice any of the quality control aspects of peer review and that peer review actually was essential to the successful operation of the press as an open access press. And if I’m honest, I think I can say that it has been incredibly influential in helping the press to be as successful as it is, both respected academically and visible numerically in terms of downloads and accesses to its published output.

Miranda Prynne: This is a, I realise, a broader issue, but do you think the problems that we are seeing in terms of certain papers not standing up to such rigorous checks or worries around a lack of reproducibility and certainly in the scientific communities? Is it ultimately the academic incentives that encourage or put a huge pressure on individuals to keep publishing are the root cause here?

Paul Ayris: Yeah, I think there is a culture. It’s to do with the research culture. And in certainly the UK, probably in Europe and indeed further afield in North America and in China, there is enormous expectation that researchers will publish. And it is still a published, yeah, still pretty much a publish or perish syndrome. And that drives publishing processes and practices and can lead to unfortunate outcomes, such as the ones you, some of the ones you, you outlined. If you’re not fully transparent about your methodology and how you reached your conclusions, if your data is not publicly available for scrutiny and not there to be reproduced to check on the validity of your assumptions, that in an open science environment, those are seen as bad practices. So, to move away, to get away from the challenges and the failings that you’ve outlined, we need to change internationally. We need to change our research culture. And open science brings with it a number of insights into new ways of working, new ways of collaborating, an emphasis on transparency and reproducibility and accountability, which will check many of those imbalances that you outlined. So, I see open science as part of the solution to a known challenge that research and researchers have, and one that we should fast forward in terms of the solution to tackle those agendas.

Miranda Prynne: Yes, and it’s important, the point you made that this needs to be done internationally. Moving on, no one can ignore the impact of AI, specifically gen AI, on higher education across the board, really, but certainly on publishing. I wondered how the UCL Press team as a whole is approaching this and managing the risks, but also arguably opportunities it offers.

Paul Ayris: Well, as a press, we’re working with other presses, commercial presses, and open access presses around the world to look at a number of issues which we all share as publishers. We’re looking at the legal challenges that have been brought against some AI companies that could determine whether or not their activities in training large language models on large data sets is or is not an infringement of copyright and so legal or illegal. That’s a really important question in the publishing world. We’re also looking at technological developments that allow gen AI tools to attribute their sources. So you not only see the information, but the AI tool will tell you where that information was derived from. And that is essential for works that are published under a Creative Commons license because attribution is part of the license. And gen AI doesn’t always understand that attribution is important. On the positive side, it’s an opportunity. Yes, I’m looking at ways in which AI can work in a publishing environment, perhaps by producing high-quality translations of press outputs into languages where otherwise it would be expensive and time consuming to produce translations. We do produce some translations, but not routinely. Is AI a way of doing that? If in China, say, not everyone will be fluent in English, can we translate our outputs automatically or on request into Mandarin Chinese or Cantonese? That’s an important question for us if we want to maintain our visibility on a global stage. Open access often supposes that English is the default language. Now, that’s not very equitable. So there needs to be, we need to look at ways in which the translation issue is brought into the forefront of our thinking. And we’re beginning to look at AI, see whether that is a way forward.

Miranda Prynne: Yes, it’s a complex thing, AI. Fascinating but complex. The doorways it opens feel endless at the moment. Some good, some bad. But I think everyone’s sort of grappling with it in their own way. Finally, I’m sure many listening would be very interested to hear from you what advice you would offer for academics, particularly early career academics, really looking to maximise their chances of getting published. What would you say to them?

Paul Ayris: Gosh, right. Okay, well, if I said one thing to them, I think it would be this: let’s take monographs as the basis. The key, I think, the single most important thing that an early career researcher or academic needs to do is to develop a really strong proposal for the book they want to publish. A proposal that clearly situates their work in the wider landscape of the existing literature and sets out clearly the value of their research and the potential audiences who would be interested in reading those outputs. How does your research make a difference? And what does it say to the challenges that face society, equality, justice, health, global warming? Does it have any impact in any of those big societal questions? Now, your proposal needs to do all those things. If your proposal can address all those challenges and questions, then you’re much more likely to get published because when it goes out to peer review, you will have ticked all the boxes that the reviewer could potentially be looking at in terms of making a judgment on whether this is or is not a publishable piece of work. And it’s not always obvious from some proposals that come in that all these issues have been tackled or considered. So I think the proposal really sets the framework for future activity when you’re writing, and is a good proposal is much more likely to be published than one that doesn’t meet all those criteria.

Miranda Prynne: Well, that’s a great piece of advice. I can see by setting things in the context, broader context, both in terms of academia, but also the real world in which we all live and work, that is also what demonstrates the unique nature of the work in question and gives it its raison d’être.

Paul Ayris: Yeah, I mean, the role of universities is both to undertake research, to push forward the boundaries of our understanding and knowledge in any particular area, but also to answer and support society in terms of answering the questions that society has about its existence, about its sustainability, and how to improve conditions in society across the world in terms of health and justice and dealing with threats such as global warming.

Miranda Prynne: Well, Paul, thank you so much for taking the time to talk to us today. There’s a lot there for anyone listening to unpack.

Paul Ayris: My pleasure. Thanks very much for the opportunity to talk.

Miranda Prynne: Alongside the publishing insights, there were some useful takeaways there for academics wishing to get published. And Paul’s final point that scholars should try to situate their work within the context of known global challenges is a really good reminder of that core mission of universities and academia to serve those wider societal needs. Because I imagine when one has spent months, even years completely immersed in a very detailed, intricate research project, it might be easy to forget to step back and take that broader view. Now, moving from a relatively well-established publisher to a new player in the game, Philipp Koellinger set up DeSci Labs in 2021 to answer a need for greater transparency and reproducibility across scientific publishing. His aim was to create a platform that would make it easier for researchers to share their work early and that would recognize the value of peer reviewers. But he can explain all this a lot better than me. So over to Philipp.

Miranda Prynne: Hi, Philipp. Great to meet you.

Philipp Koellinger: Yeah, great to meet you, Miranda. Thanks for having me.

Miranda Prynne: Pleasure. So you are the co-founder of DeSci Labs and co-creator of DeSci Nodes, both of which aim to support a more transparent, accessible, fairer system for sharing science research. Is that a fair summary?

Philipp Koellinger: Oh, yeah, absolutely. You’re spot on with all of that.

Miranda Prynne: So can you talk us through how these platforms, I mean, I’ve mentioned DeSci Labs and DeSci Nodes. Can you just briefly explain what they offer, what they do?

Philipp Koellinger: Yeah. So DeSci Labs is the company, and then DeSci Publish is basically the first product that we’re making available to the scientific community. And what this is, is basically a pre-print network on steroids. So it’s a place where you can share your work in progress and also your past publications. So not just the articles, but also data, code, images, anything that supports your research finding and documents how you ended up with your results. And in contrast to traditional publishing solutions, everything on DeSci Publish is versionable. So you can go in any time, update it easily without overriding the original content. And we provide a number of nice AI-powered tools that give the authors additional information about the work and basically helps them streamline the publishing process. So that’s a short summary. And what will come next is basically journals on DeSci Publish. And with the journals, we will launch a reward system for referees that will actually incentivize faster and higher quality peer review services that are going to be publicly available. So again, in an effort to increase the transparency and the quality of the research that gets shared.

Miranda Prynne: Brilliant. That’d be great to come on to talk a bit more about that peer review system. That is an area of contention. I’m keen to just, for now, though, stick with the DeSci name. I gather this stands for decentralized science. And I wanted you to, again, just explain what this means and the implications of it.

Philipp Koellinger: Yeah, so you’re right. So, decentralized science basically uses a variety of peer-to-peer technologies, such as blockchains, but also more broadly, open source software that can be used on peer-to-peer networks in an attempt to make research more open, more collaborative, and more accessible with the goal to accelerate scientific progress. And there is a number of projects out there that are tackling basically different parts of the problem sphere that you have in science. So, there are several projects out there that focus on specific diseases or specific problems where a scientific solution is needed. They often emphasize new funding models to actually fund the research that is necessary to make progress. And then you have solutions such as ResearchHub and what we’re building at DeSci Labs that are more concerned with the publishing aspect of science.

Miranda Prynne: Yeah, and it sounds like one of the approaches here is very much that you’re focusing on publishing the whole process rather than simply the outcome.

Philipp Koellinger: Exactly.

Miranda Prynne: Okay. Now, I know in your day-to-day, you’re also a professor in social science genetics. So, clearly, you have an interest through your academic work in the world of academic publishing and in advancing science. But a lot of people in academia have those interests and they don’t all go on to found R&D tech startups. So what led to the founding of this? What spurred it?

Philipp Koellinger: Yeah, so I mean, this is obviously a longer gestation story that is behind this. But ultimately, what happened was that as I started my academic career, I started branching into a number of interdisciplinary research topics. So originally, I was doing research on innovation and entrepreneurship, and then very quickly ended up in a niche of the literature where people were doing candidate gene studies and other biologically informed study designs. And it was just clear that these studies are statistically completely underpowered, that they’re very unlikely to replicate. And yeah, so we tried to replicate many of these results that ended up getting published in very highly cited journals and where entire books were written about them. And we couldn’t replicate any of these results. And we also couldn’t find anybody who would be willing to basically publish replication studies that would show that original findings that were based on much smaller sample sizes actually do not replicate. And I think this was sort of like the spark that ultimately led to this growing frustration about many of these practices that we have in science and in scientific publishing. Where, you know, very often journals emphasize much more publishing papers that make big claims about novel results rather than publishing studies that show that earlier results do not replicate. And the result is, of course, that we have many areas of science where scientific progress actually ended up sort of like in a cul-de-sac where, you know, a lot of follow-up work is based on things that are ultimately not true or not robust. And this is slowing down scientific progress. It’s sort of like the self-correction mechanism that we need to have in science is clearly not working very well. And I think this was the starting point for me to start thinking about how we could solve that. And yeah, it was just a complete coincidence that I ran into people that shared the same, you know, the same sentiment and that actually had some ideas about how we could go about this. And the irony of the entire thing is, so as I said, so at the beginning of my scientific career, I was actually researching why people start businesses and like from a purely economic perspective. So, I’m an economist by training, from a purely economic perspective, it doesn’t really seem to make a lot of sense because you end up spending a lot more time, you earn less, you have this huge risk of failure. It’s harder to find a job elsewhere after you failed with the business, right? So, there’s like all these things, and so we’re trying to figure out what is going on. And the number one reason seems to be that entrepreneurs are overconfident. And now, the irony is that 20 years later, I’m one of these overconfident people. So, here we are.

Miranda Prynne: Well, at least you’ll be doing it with your eyes open.

Philipp Koellinger: Absolutely. So, overconfident and self-aware is probably a different thing.

Miranda Prynne: I mean, what you’ve described there about the motivations behind why you wanted to do this and the frustrations you’re feeling, what you’ve described is an incredibly inefficient system, isn’t it? There’s a real parallel one can draw with news and newspapers where they are only interested in a headline-grabbing story and not actually perhaps something of real kind of wider public importance that might not just grab people’s attention in that same way. So, I guess that is one problem that one could point to with the model of a commercial academic publishing sector. But that’s not the only issue here. What do you really see as being the main problems with academic publishing?

Philipp Koellinger: So let me start by saying that, you know, I just like basically everybody else in the scientific community am frequently publishing in these traditional journals and very often I’m very proud when an article finally gets accepted and it appears in these highly prestigious journals. So there is certainly there is a draw, there is an attraction and they do serve an important need in the scientific ecosystem in terms of offering these services where the content that scientists are producing is getting reviewed and curated by topic and according to particular theme. So I think all of that is valuable. But there is a number of problems that are related both to the technology that is being used and to the business models that are basically driving these companies. So in terms of business models, there is basically only two ways at the moment how you can publish an academic article. So either you put it behind a paywall, which means that you give up the copyrights and the publisher then owns the copyrights. And universities then have to pay enormous subscription fees to get access to the journals where their own scientists published in. And this is problematic for the obvious reason that these high paywalls where very often you need to pay more than $30 or $50 to get access to just one particular article. This basically locks out the vast majority of mankind from the actual scientific publications. And it also leads to this bizarre situation that sometimes, as an author, you don’t have access to the paper that you publish because the university doesn’t have a subscription to it anymore, right? So this actually happens several times. So this is one possibility how they finance themselves. And the other possibility is with so-called open access publications. And there, the author has to pay a publication charge. And this is only paid once an article is accepted. And these publication charges can range from $1,000 at the lower end to up to $12,000 or $13,000 in highly prestigious journals. And then the papers are publicly available, which in principle is a fantastic thing. But this comes with a perverse incentive for the publisher to lower the acceptance standards because every time they reject a paper, they are actually losing a substantial amount of revenue. And what we’ve seen is that in the past 20 years or so, since open access really started to take off and to become more and more popular and dominant, we’ve seen that the overall number of publications has been like exponentially increasing. So there’s just more and more and more getting published. So it’s like over 3 million articles every year that get published at this point. And a very large share of these openly accessible publications are actually published in journals of very dubious quality, right? So there’s, you know, we’re talking about paper mills, we’re talking about predatory publishers. There have been so many scandals related to this. And of course, the problem is like, if you as a scientist can basically pay somebody to publish your work, how can you ensure that, you know, that there is going to be an unbiased, objective, and thorough evaluation of the quality of the content that you’re releasing? So these are the basic two problems. And then there is a third one on top of it, which is that the referees, which are typically, you know, other members of the scientific community, so peers that you may or may not know, but that basically work in the same area as you. So they’re doing these peer reviews for you, but they’re doing that without getting paid or getting any sort of other rewards for it, right? So basically, the scientific community now, we’re not only doing the evaluation and the curation of the scientific content for the publishers, we’re not getting anything paid for producing the content or for doing these services. And then we pay to either get our work published or have access to it. So, this entire system is just, it’s really absurd when you think about it. Yeah, so this is like on the business side. And then on the technology side, there is the obvious problem that most publishers basically only have an infrastructure to publish articles. So, basically, papers, right? So, PDF documents. And although this is very valuable and people really like to read scientific articles and they’re always looking for the option to just download the PDF, it also means that a large part of the research and of the evidence that the claims in the article are based upon is not available, cannot be evaluated, right? So, papers typically are not published with the data or the code that are underlying the results, which means that they cannot be evaluated. Others cannot check if these results are actually true, what the authors really did to end up with their claims. So, it boils down to having to believe what the authors are writing. And as you can imagine, this leads also to a bunch of problems.

Miranda Prynne: Yes. And unfortunately, you cannot just simply accept the claims put in these papers. I think it’s important that you stated that the big academic journals do play a really important role. A lot of people could come under the impression that a new project would be seeking to almost replace them. And that doesn’t appear to be what you’re trying to do at all. What you have pointed out is there are some big gaps in the way it works at the moment and need filling. Talking then through the business model, this business model that you described is clearly just a completely one-way relationship with perverse incentives and all sorts of other problems. So how can these stumbling blocks be addressed?

Philipp Koellinger: So I think that there is several ways how the system can be improved. And just to also say that upfront, so we’re not the first ones who thought of these potential solutions. So many of these ideas have been around for a while and people started experimenting with them in some way. So one of the ideas here is that instead of doing all the research first behind closed doors and then only making the final product, this article, publicly available in one way or another, where also the entire process of how this was evaluated and why it was published or rejected is entirely hidden in a black box. So, rather than this, how about if we switch to a model where we’re actually working in the open, where we’re doing research in the open, and where we do the publication first, and then we do the review and the curation of the publicly available content later? And there are other publishers that have actually tried to experiment with this. So, eLife, F1000, there are examples of this. And in principle, I think this idea of switching the paradigm around and like publishing first and then doing reviews and curation later is very attractive because it, first of all, speeds up the availability of new research results. Just sharing your manuscript openly on the internet is actually not a cost at all, right? So, this is like this should be for free and it is for free. So, there is no economic reason or technological reason that stops people from doing that. And then, to do the review and the curation of that publicly available content in the open still fulfills the same purposes as the current journals do, but it just creates much more transparency around the entire process of how the authors ended up with their results and why a particular manuscript ended up getting accepted in one journal and not another. So, this is one thing.

Miranda Prynne: It also introduces much greater accountability. The whole process, why something’s been accepted. But the other thing that strikes me, hearing you describe that as a flipped approach, is that it also is a better reflection of the scientific process. You put forward a theory, you test that theory, you refine, and so on. That’s much more reflective of how it works.

Philipp Koellinger: Yeah, absolutely. And I think the number one reason why people are not doing it, so why scientists are not comfortable working in the open is that they’re often afraid that their ideas are going to get scooped by somebody else who then ends up publishing the same thing basically before them, right? And this is based on this idea that only a journal publication actually counts and everything else is irrelevant. So I think actually this is not true. I think the other way, rephrasing that and thinking about when you share your ideas very early on and you claim provenance that you had this idea at a particular point in time, then you already claim credit for that. And you can always point back to that and you can say, look, I was first. So from my point of view, this fear of getting scooped is actually not a legitimate reason for for trying to work behind closed doors. I’m a big fan of much more transparency.

Miranda Prynne: Yeah, absolutely. The earlier you get your idea out there, even if others are working on it, you are then part of that process and part of that thinking. I guess another point of contention among some in or among many in academic publishing at the moment is the emergence of gen AI. And there’s a lot of controversy around gen AI being allowed to access and crawl through huge databases of scientific journals with authors saying, well, hang on a minute, you know, should this be allowed? Isn’t this our work? Isn’t this an infringement of intellectual property and so on? What are your thoughts on this? Should these AI tools be able to use the data? Should there be greater accreditation or some sort of two-way reward system that acknowledges the intrinsic value of this stuff?

Philipp Koellinger: I mean, this is a good question. So honestly, I don’t feel qualified to take a stand here in terms of the legal implications of copyrights. But as a practitioner, as a scientist, I can tell you that I would much rather work with the help of AI tools than without them. So they offer a lot of possibilities of just speeding up the research process, being more efficient. They help clean up the language when you’re writing an article. So when you’re using these AI tools, of course, you have to say, like, for which particular reason or for which particular purpose you use them and just be transparent about it. And as long as you’re doing that, I think that that’s totally okay. But they’re very, very powerful in terms of our ability to speed up scientific progress. And I would rather have large language models that are trained on the scientific literature and that take that into account than models that are only trained on Twitter content.

Miranda Prynne: Yeah, so the benefits outweigh the negatives from your perspective. And that’s a good argument that these things aren’t going away. We want them to be well informed. On an individual level, coming back to the DeSci project, what opportunities does this offer? Can you talk us through sort of on a practical level how someone might benefit from what you’re offering?

Philipp Koellinger: Yeah, absolutely. So we’ve built our tools really with the objective to make it as easy and rewarding as possible for scientists to follow best open science practices, which means being transparent, sharing your researches, working in the open, and really starting to share your ideas and your progress very early on. And it’s super easy actually to do that in our system. So we have a variety of AI tools that make it very easy for people to quickly share their results. So instead of like battling with a submission system for hours in our system, it’s like one minute if you already have a manuscript and then you’re done with your publication because all the metadata is automatically extracted. But more importantly, you can actually use the system to build this credible track record that shows the public, the reviewers, your colleagues, and editors the steps that you took to arrive at your final results. And it turns out that this type of transparency is actually extremely highly valued by editors and by referees. So by doing that and by also sharing your data and your code, actually you are improving your chances of getting published in the journals that ultimately count for your career. And you ultimately also increase the potential impact of your work. So there’s plenty of studies showing that articles that come with data and code and other resources, that they get cited much more often, right? So you can just say that from a very utility maximizing perspective of a scientist who works in the established system, this is actually a way how you can save time and basically get more impact out of the work that you’re doing anyway. And then we have a number of nice analytics tools that are actually quite helpful for scientists. So a couple of months ago, we released so-called novelty scores. So they’re basically the result of a vast model that was trained on over 250 million academic articles and that measures how surprising the content and the context of a particular article is compared to all the existing literature out there. And this is nice because we made these scores available now for really all these 250 million articles that can be accessed either as a full text article or at least where you have access to the abstracts through OpenAlex. And so with these novelty scores, you can actually find very quickly the most groundbreaking ideas, the most challenging new things that are most likely to move the needle. And you can use the search engine to filter by particular topics or universities or authors. And you can also get all these analytics for every version of every article that you upload on the platform. And then you can make the decision when you want to submit your article to a journal. You can actually either reveal these analytics and basically create a cover page for your submission that reports all these statistics and that also shows whether you, for example, share data and code and whether others have already acknowledged your work. And all of that may actually help the authors to just be more successful in the traditional academic ecosystem. So this is the value that we can provide in the short term. In the long term, we will be able to also enable authors to just get on-demand peer review and to onboard journals into our ecosystem that can then also reward their editors and their reviewers for doing fast and high quality work. And there, again, the idea is making everything much more transparent and then also giving credit to the people that actually do the work.

Miranda Prynne: So one thing you’ve described there, the analytics, it’s really helpful, as you say, for then submitting your pieces, writing pitches for academic journals, but for just demonstrating the impact. I mean, helping you get, in a way, more bang for your buck, because you’re able to demonstrate impact at each stage of your research rather than just the final bit. Okay, great. You mentioned at the start of our conversation, and we’ve just come back there now, bringing in journals into your system. Is that the next stage that you’re working towards at the moment?

Philipp Koellinger: Yeah, so really the idea is that, I mean, everything that we’re building is open source technology and like all the content is shared on this distributed peer-to-peer network. So, there is no data silos like in the traditional system. So, there’s also no possibility to create paywalls around that content. And instead, so what we’re going to do in our system is what people typically describe as diamond open access. So, these are journals that are neither charging authors publication charges, nor do they lock up the content behind paywalls. And they will be able to not only publish papers, but also data and code and all the other artifacts. And then, the idea really is that we will have this token reward system that will enable the editors to actually be much more successful in finding referees for evaluating the work that gets submitted to these journals. And then these referees will also have an incentive to actually do much, much faster and high-quality work. So, we actually think that the current turnaround times of a journal, which are on average, something between three to six months until you hear back from the journal whether you got rejected or whether you’re invited for a resubmission. So, this is how long it takes for the first iteration. So, we think that we can probably bring that down to two or three weeks without sacrificing quality and instead just creating much more transparency.

Miranda Prynne: Wow, that would be very welcome, I think. So, essentially, you’re providing a tool for the journals to use. Is that the idea?

Philipp Koellinger: Yeah, yeah, exactly. And the idea behind this token is really that we want to create a system that basically shifts the monetary value that scientists are creating with their referee services, which currently basically boil down to a multi-billion dollar donation that science makes to the for-profit publishers every single year. So, we basically want to have a vehicle that allows them to capture the value that they’re creating themselves and basically give it back to the scientific community. So, these tokens will be freely distributed to bona fide scientists and to journal editors and people that want to use the system to evaluate and curate research. So everybody will be given basically enough of these tokens so that they can operate in the system for free as long as they keep behaving like a good citizen, where they give back at least as much to the community as they request from the community in terms of getting their own work evaluated.

Miranda Prynne: Yeah, okay. So it’s self-sustaining, a circular system. I think the time efficiency element on an individual academic level, the fact that this can really facilitate an overworked academic to get their data in there quickly will be very welcome indeed. One final question, which is we’ve been very, very focused on science. Is there a potential for this opening out to other disciplines?

Philipp Koellinger: Right. So what we’re building is really focused on scientific publishing and making scientific knowledge more accessible and more transparent. But I think there is very obvious use cases for what we’ve already built outside of traditional academia. So with these novelty scores in the search engine that we have that basically gives you access to like the entire scientific literature, you can already use the system to very efficiently find new research that you’re interested in or that’s potentially relevant for your company or for the research process that these companies are doing. And it’s also an interesting way for students to just familiarize themselves with the newest trends in particular fields of science. So all of these things are already possible.

Miranda Prynne: Well, it’s an amazing resource and I think DeSci is forging the way probably for many others to potentially try and do similar things, as you say, in other areas. So watch this space. Thanks so much for talking to us today. It’s been absolutely fascinating.

Philipp Koellinger: Thanks so much for having me, Miranda. Great pleasure.

Miranda Prynne: So if you started this podcast feeling that academic publishing was doomed, I hope these two interviews have given you renewed hope for the future of research and the possibilities and opportunities currently being explored. There are a lot of passionate people out there seeking new and better ways to advance knowledge and Paul and Philipp are two great examples.

Miranda Prynne: You can also find a Campus spotlight guide looking at what you need to know in academic publishing today at www.timeshighereducation.com/campus. This pulls together advice from academics and publishers all over the world on pitching your ideas, perfecting your manuscript, understanding open access and even launching your own journal.

Miranda Prynne: If you enjoyed the podcast, please like and share, and we will be back with you in a couple of weeks. Bye-bye.

Sound clip: You’re listening to a Times Higher Education podcast.