Cadmore media player playing video The Diminishing Importance of the Article: An Overview of the Emerging Modular, Multi-modal Research Landscape
Video Player
Accessibility
Using forms mode, you can utilize keyboard shortcuts to make usability more efficient.
Space bar will play and pause the video. Pressing escape will close all displays and pressing it again will jump to the end of the player.
You can press A to open the Keyboard Shortcut Dialog which also provides an audio description of the player and all of its features.
Language
1x
Play Speed
Adjust Volume
Transcript
Search
Result Count
of
Click to jump to result
Search
Re-center
Segments
Result Count
of
Click to jump to result
Search
Name:
The Diminishing Importance of the Article: An Overview of the Emerging Modular, Multi-modal Research Landscape
Description:
The Diminishing Importance of the Article: An Overview of the Emerging Modular, Multi-modal Research Landscape
Thumbnail URL:
https://cadmoremediastorage.blob.core.windows.net/5044f55c-2d77-4411-aa60-c8215f642475/videoscrubberimages/Scrubber_1.jpg
Duration:
T01H02M03S
Embed URL:
https://stream.cadmore.media/player/5044f55c-2d77-4411-aa60-c8215f642475
Content URL:
https://cadmoreoriginalmedia.blob.core.windows.net/5044f55c-2d77-4411-aa60-c8215f642475/session_4e__the_diminishing_importance_of_the_article__an_ov.mp4?sv=2019-02-02&sr=c&sig=ajeB%2BQ1sSMXCpbDJYgrfVrwv5y3Jrsjh0Y0uf0%2BcFpM%3D&st=2025-04-29T19%3A44%3A25Z&se=2025-04-29T21%3A49%3A25Z&sp=r
Upload Date:
2024-12-03T00:00:00.0000000
Transcript:
Language: EN.
Segment:0 .
Hi, folks. I'm Bill casdorph. I'm pleased to be moderating this session on beyond the article. I've done a number of presentations or written a number of things recently about the session originally called it is still called the diminishing importance of the article.
And I want to start out by emphasizing why. Why is this not advancing. Technical person still in the hallway there. A minute. Yeah that is. I got it. So, no, I'm not saying that the scholarly article is going to go away.
This is not about the death of the article. Article is indispensable to scholarly publishing. And just that it's been around for 350 years. Here's the first one you probably already know about that. Philosophical transactions 1665 still in publication. This is the world's longest continually published journal. And the article is just the gold standard for publishing scientists.
Scholarship has been free for decades and decades and still is to this day. This is the Watson and Crick paper announcing the announcement of discovery of DNA. My favorite spot in Cambridge. I did a lot of consulting with Cambridge University Press and I would always go to the pub where. He announced this discovery to his colleagues. And I sit-in the very chair that he was sitting in to do that.
I think that's pretty cool. But today, the article, I guess I would say is maybe necessary. I think Kristen may argue with that. She's the last speaker, but not sufficient because we need many, many more modes to communicate scholarship. This is really about scholars in general, but a lot of it is really centered on science. The key principle is the need to communicate about the research.
Throughout the research process, not just a year or two after the project is done and an article gets published. What are the issues. Well, first of all, as I was just saying, the article often isn't published until a year, maybe two years after the research is finished. And one thing that has been bugging me my entire career in scholarly publishing 40 some years is at negative results often don't get published.
And that's ridiculous because that's really important. Somebody else is going to duplicate that investigation. It's just important that got published, but it often is not published. The other thing that is commonplace nowadays is people realize that the research data needs to be available and findable. And we'll have Tim talking about that, or at least Tim's in the business of helping facilitate that, I'll put it that way.
But if you're really going to replicate the research, just the data isn't sufficient. You need code models, other resources. And ideally the whole concept here is that this sharing of this information and these resources needs to be done along the way as the research is underway and the text just isn't always the best way to convey some aspects of that. So just to reiterate, no, I'm not saying the article is going to be obsolete.
The formerly published peer reviewed article is still the gold standard in most disciplines. Notice I said, most not all. And progress in addressing the issues is being made. So here's just a quick run through of things that the speakers following me will get into some of these topics in more detail. But in terms of publishing the research.
Along the way and certainly not a year or two after it's done is obviously the growth of preprints and preprint servers. That's become very common. Some fields like math and physical sciences, they're used exclusively and oftentimes there is no actual final peer reviewed article. It's those practitioners, those researchers depend on preprints to communicate their research. And we've got a very good example of that.
I'll get to that in a second. And another thing that people sometimes kind of forget about is that there are disciplines where it's really conference papers that are the. Substitute in terms of validation and promotion in fields like engineering. More important than published journal articles, it's the papers that they live and die by in terms of negative results.
That's something that preprints can do. If some of you saw the session yesterday that's quite related to this new model from eLife is, I think, a really interesting thing where there's open peer review and the author gets an opportunity to revise. They don't have to revise it. They can just publish it that way. It lives, makes no accept or reject decision. The author determines whether they want to.
Respond to what the reviewers have said or not and whether they want to send it to a different journal or whether they want to put it up on eLife. And then. We all know that there are lots of data repositories. The difficulty there is in those three little words. Words? lots of two. Two words. Lots of data repositories, because it's kind of a mess right now in terms of where all this data is and how you can get to it and how you can find it.
So dataseer is an AI based system for helping determine what data to share, where to share it, how to share it. And that's Tim's brainchild. So he'll be talking to us in a few minutes. And in terms of models, Code Ocean has been around for quite a while. I think it was I think it was the winner of an ASP preview session about five years ago or something like that, I recall.
And that facilitates actually sharing the code and the computational research between teams. So it's a very useful thing. So my point here is these things are being done. This is not speculative or things like here's what we should be doing and aren't. It's like, no, we can and are doing these things just not as extensively as it should be done. There's a new format that I had the pleasure of co-chairing the Nasa working group that developed the CPLD standard.
It's a standard that's really designed for this because it enables packaging of arbitrary portions of content along with data and semantics and other resources into a single web based format that's optimized for interchange, search and display. And one thing that I think is really important about this CPLD format, it's quite new, so it's Elsevier uses it extensively, but it's not. It's just early days for this to roll out through the research ecosystem, but it's entirely based on web standards that everybody already uses.
So it's Schema.org, json-ld, some other W3C standards that may be less well known, like annotations and application manifest, et cetera. But it's entirely W3C standard based format. And then obviously, for different modes other than text for communicating research videos, 3D molecules, LiDAR, an archaeological dig, an interview with a famous poet, a documentary of a World War two battle.
Just to point out that this is not all only about science. So really the fundamental thing about scholarship is it's collaborative, and that's how we make progress. So I want to introduce I'm just going to briefly tell you who the speakers are going to be. And in what order, and I'll let them fill in any biographical details they would like to mention about themselves. But we're going to start with Stephanie orphan, who's the program director at archive.
And I think I'll leave it for Stephanie to tell you how long archive has been around so that you don't think this is a cool new thing. Well, it's a cool old thing is what it is after Stephanie to really demonstrate that this isn't just about science is Susan from the University of Minnesota Press. They've got a really interesting interactive platform called manifold that she'll be able to talk about and actually show, ideally, assuming we've got good connectivity that really enables that kind of dialogue and interactivity between people involved in a research endeavor, followed by Tim, who I've mentioned is founder and CEO of dataseer, which is this AI based service.
And I'm not sure that he's necessarily going to talk about dataseer. He's free to talk about whatever he wants to talk about. And he's shaking his head that, no, he's not presenting about dataseer, but you just need to know that that's who he is and. But apparently God was pleased about that.
And then finally, I will say the person that has really. Open my eyes to all of this for some time is Christina tan and she'll be bringing up the wrap up here. She's the founder of stratus, which is stratus strategies for open science. And she's deeply involved in this transition from. Static text based publication to dynamic ongoing publication.
Each of these presentations has got to be ideally quite short because we want to have plenty of time for Q&A at the end. So I will hand it over now to Stephanie. And I hope you know. You are not. You don't have slides.
I don't have slides. So bear with me on that. I have some very brief initial remarks and also very much listening to what my co-panelists have to say and how the conversation evolves after that. So I'm just going to talk a little bit about archive and preprints and I will start with that. So I think it's safe to say that archive was an early disruptor of the traditional scholarly publishing process.
For some background, back in 1989, Joanne Cohn, who was a physicist at the Institute for Advanced Study, had started an email list through which she was distributing scholarly articles in high energy physics or string theory, I believe, that were marked up in the tech format that really took off in the size of the list grew to such a size. I think it was like 180 people and she was individually handling requests.
It became untenable to continue on in that fashion. And a young researcher at Los Alamos National laboratory, Paul Ginsburg, offered to automate her email list. And that was essentially the birth of archive. So archive stayed at Los Alamos National lab for about 10 years until 2001. At that time, Paul moved over to Cornell University and he brought archive with it.
So for many years, archive was housed at the Cornell University Library. Many people still think it has a relationship with the Cornell University Library, but it does not. At about 2019, it moved over to the Computer and Information Science department and very quickly after that, by 2020, it migrated over to Cornell Tech, which is a graduate campus of Cornell University in New York City that's focused on primarily computer science.
And it does kind of like a marriage of business and computer science and things like that. It ended up there because the person that had been the Dean at kiss moved over to become the Dean at Cornell Tech, and he's a big champion of archive. So that's the history of archive and archive was and continues to be continues to be focused on the article is a research output.
But the introduction of preprints greatly changed the game. It allowed for fast dissemination of research so authors can get their research out there. Other people have access to it and can build on it. Very importantly, especially in the hard scientists facilitated claim staking. So a lot of people get that time stamp on archive and they can say, see, I thought at first I put it on arXiv, but over the last 30 plus years, things have changed.
Archive has grown significantly. We now have over 2.3 million articles. That's just like the number for like primary articles. In addition to that, there's probably double that in updates and versions and things of that nature. 60% of the content on archive is also published in traditional journals or through conference proceedings, but 40% of it isn't found anywhere other than on archive.
As I said, a lot of things have changed over the last 30 plus years since archive was born. I don't know if you've noticed, but preprints are having a little bit of a moment in the last couple of years. There's a growing number of preprint servers in a number of disciplines and some of them are discipline agnostic and there's a lot of experimentation currently going on around mechanisms for peer review as a basis, capturing comments using preprints as a basis for content for journals such as an overlay journals.
So that's kind of my little setup for what's going on with preprints and how we started this little shake up. But preprints aside, there's a lot of innovation and thinking going on in scholarly communications in general, and the other panelists are going to talk about that. And I'm interested in hearing what they're going to say. It's a really exciting time.
I have to stand on my tippy toes to see you all. I will do that. Sorry so I'm going to talk about the humanities and the social science perspective, especially from a University press. And there are others in the room that are doing similar things to me. I can see them out there basically at the University of Minnesota Press in about 2010, we were publishing a book called debates in the digital humanities, and we partnered with Matt Gould at the City University of New York for that.
And he wanted to do open peer review, and he wanted it to be a website that was accessible during the peer review process where everyone could read and comment and incorporate that work. And we created a website with him with an agency called cast iron coding. And the reason I'm telling you this story is because what we built for that project, that one book, eventually evolved into what is now manifold, manifold scholarship.
At the University Press, we had more and more scholars coming to us. Many of them digital humanities scholars, but not all of them. And they wanted to do things that were a little bit different than what they could do in a traditional print book. They wanted to publish in a way that was iterative. They wanted to publish as they were publishing or as they were creating the work. And I'm going to go into an example here.
Cut, copy, paste to show you an example of this. So what this scholar Whitney Turchin did with this book is that she started with resources. She had these collections of media and audio files, and she this is where her research began. So when we were building manifold, one of the things that we thought is what is involved in a book. And what we saw was that there were texts, there were media of all types.
And of course in the print book all you had was a static image. But in the web we have the opportunity to do lots of different things, including audio files data sets. And what she wanted to start with was an example of this. So this project for the first two years was just a collection of resources. And then the next thing that she was able to do, she was 41 total collections of resources here was that she started to write her book and we would work with her to clean up the text in a minimal way and put the draft chapters online.
And so you could follow the evolution of her research and her thinking as she was developing this book, because this was a new idea. She created this section on how to read this book for the reader, to orient them in what they should be expecting from her project as it was being developed. And eventually then of course, it became the full text that we know as the book.
And there is a print, there is a print piece that goes along with this. We have an actual book to go with this. So every manifold book we bring in an EPUB file and that becomes the base text. And then within that base text what we can do is annotate it with the resources and the media collection. So that's what you see in these side sidebar is these are where the media and the resources within her project fit within the text.
And then you can but we also have the original image from within the PDF and the EPUB file itself, and it's created a much more dynamic interactive experience. In reading the book. You can learn a little bit more about the different videos you can watch while you're reading if you want. You don't have to do that. We offer annotation and some different reader tools here and what we've done really, we think is given a more dynamic reading experience for.
Sorry It's hard to control and talk at the same time. What we've done is we can. We can create reading groups for classes. You, as a reader can annotate and highlight as you're reading, if you're logged into the manifold site. And we've created really an interactive reading experience in a different way than within the print book, classes you'll see have dialogues and interaction within the book itself.
So that there's some discussion outside of the scope of the class. And it has become a nice really. Tool for both for reading but for education. But then after we. Started with the book, then we decided to bring shorts series into it. So forerunners is a series of 35,000 words or less essays. Many of these might evolve into a full monograph, but this is another example of a semi formal publication that is an idea.
It's one idea that will eventually perhaps turn into other books and outcomes. But we allow our writers then a forum to have that discussion with their readers online. We've also brought in journals and. Which you see down here. Critical ethnic studies was our first journal that we put on the website. So in this way, what we do with these projects is we have a series of content blocks.
And so you can have texts, you can have syllabus, you can have media resources, and it's just a much more dynamic and multi-modal. Each of these things can be published as they are written rather than all at once as a final publication. So you can watch a work be developed and built as it's being published and written. And we've got about 30 different universities using it now. Cuny is using it for their whole system where they're publishing OERs where publishing papers where publishing class projects.
They have about 800 projects at this point on their manifold website. We also see Brown University using it for their campus wide readings. So they're bringing in a mix of texts and discussion and videos, panels with their scholars on campus and their entire campus. Then, as you can see with these annotations, are commenting and interacting as a community around publication.
It's one mode that we think addresses the need for the Humanities and social science, and we're excited to see what others are doing with their dynamic publishing. And of course, we'll continue to evolve manifold as we go. And now I'm going to turn it over to Tim. So we're coming back, back, back. Of course.
And just like the coyote. The Did we have a PowerPoint up there earlier.
OK, that's it. This one. Sweet and then we do this, I believe. But Thank you, Bill, for inviting me. So in contrast to the beautiful, well-developed ideas that have become products that have in some cases been in use for many years, this idea what I'm presenting is about as half baked as it is possible to be. This is an idea.
It's a concept. I think it sort of dovetails quite well with your CPLD. So why is this. So I've enlisted the help of Wile E Coyote. We're going to use Wile E Coyote to blow up the article and we're going to see what bits emerge and if any of those bits are of any use to anybody.
So we've blown up this Article I published by me in current biology, and here are the words I could have done letters, individual letters. But this is the bag of words, the first 150 or so of the 16 or 30,000 words in this article. And they're no use, really. I suppose you could plot a graph of how many there are of each, but it's not helpful.
If we use slightly less TNT, we might end up with sentences. Now, these are useful because. They tell us things. They contain subjects and objects, predicates and but there's no context. We're not sure whether these two sentences here. A related to each other and how they relate to each other. One is a sort of a statement of what we're going to do.
And then the next is the result of doing another thing. So if we take away even more TNT, we end up with blowing up the article intersections. And I'm going to argue that this point we actually get to something useful. And so this is a paragraph describing the application of a statistical test to a question and the result of that statistical test.
And we also find a graph in the article somewhere else in the smithereens. And we also find in the article a link to a data set and that data set is online on Dryad. And part of that data set corresponds to what we're talking about in the article text. So we have year, and whether or not the data were shared. And that corresponds to what's being tested here.
So now we're starting to put things together. And this reminds me very much of a semantic triple. That is we've put. We've got a sort of a question. We've done something to it. There's a process applied. And there is an object on the other side, which is not entirely unlike this.
So we've asked the question. To what extent is whether or not the data is extant predicted by the age of the paper. And we've brought data to bear on that question. We've ended up with a result. And this triple here, semantic triple is the raw material for machine learning. So I would argue that this three way combination of the section saying what's happening and what the result is.
And the result and the data that are used to test that. All the raw material that AI is going to use to understand science. We have a question posing, asking about the relationship between two variables. We test that and then we end up with an assessment of the strength of that relationship. And that can be Fed into an AI model along with millions of other triples to teach it stuff.
And so this way of structuring articles could become very powerful way to properly get AI to understand science. Because right now, to be honest, large language models are a bit like they've moved from asking a very intoxicated person at the bar what it thinks to ask your uncle that went to University. And so he'll tell you what everybody says about it the received wisdom on something but LLMs don't understand authority.
So they don't understand that or they give you the JPEG version of what the internet says about something. And so a lot of people are wrong on the internet. I don't know if you've noticed, but and so almost impossible to pick out. This is the authoritative statement on this and therefore this is correct. Whereas these triples. Give the strength of evidence underpinning an opinion.
And that's how LLMs can be taught anyway. OK done. Thanks you get. Good boy. Yeah, we are. Fuck You're ready to roll. No, that's Tim's.
That's Tim's bar last night. Get that. It's a little weird. You're not this beyond article. This one. Yep there you go. Hey, there's a lot of fret about this. Most of my fret about whether we'd get this.
Is it. OK Yeah, let me just. There you go. Awesome Thank you, Isaac. He's a miracle worker. Appreciate Isaac.
Hello, everybody. My name is Khristian rutten. And I as Bill mentioned, I'm the founder of strategies for open science, or stratus, which is a consultancy and implementation company in open science. And I work a lot with funders, some institutions, not so many publishers, but a partner with some. And then I also have a nonprofit called incentivizing collaborative open research, or I-corps for short.
And we started I-corps to do research on open science so that we can study ourselves and learn what works and what doesn't. So the evolution of media sorry is a really silly picture. I don't know why the guys are naked, but the guy with the computer at least should have clothes. It's totally weird. But I think we're sort of here when it comes to where we are with scholarly publishing. We're kind of at this laptop guy's spot where we're delivering things on the internet and maybe a phone or two.
But our formats are really still very, very print based. We are in the world of print legacy still 20 some years after putting journals online, we still are tied to the article, this container that made sense in a magazine or a book like presentation. We are still using issues like what's that about. The journal itself is like a book that had to get on a truck. So people had to gather up all the articles, create an issue, put this thing on a truck for it to go places.
So people could read it. We've lost the trucks are not involved anymore. So like but we're holding on to these formats and these kinds of structures and it's just not the way the rest of the internet has gone. I feel like there's been some sort of self-isolation. Let's see if I can move this thing in. Nope it says self-isolation and publishing partner. OK, so I think we've been in this kind of weird world in publishing where we started out along the same lines as all the other media with print and things like film and databases.
Things went online. That was like a big moment that happened. For those of us who are old enough to remember, that was before the web. And then the web happens and things started to bifurcate. And on the publishing side of things we embraced. And I say we because I was there participating in these things, committing these sins. So we embraced XML.
We thought HTML was too flaky for the gravitas of these things that were publishing. And it was the Wild days of the web and people were doing wacky things and changing web pages like from day to day or minute to minute. And that felt a little scary when you want this content to remain in its canonical form. So everybody got behind XML and the PDF made a lot of sense. We did things like we printed tables as images.
I remember doing this like it was pretty embarrassing. Finally, we got tables to be tabular data. We added some interactivity to figures zooming in and stuff and then links to all of these other rich resources that, as Tim was pointing out, are required to really understand the work, like data and code. Meanwhile, everybody else, literally everybody else went down completely different pathways and got all these goodies right.
So all of a sudden their apps, social media, all this information is happening across all these different platforms and formats. Augmented reality, virtual reality. And so I think we've lost a lot of opportunity there by having stuck with this pathway that really came from the trucks and has moved forward into this static state we're in now. And so we're stuck in this single mode and this has actually been really solidified through the business models and publishing.
And regardless of whether it's subscription or whatnot, it's still an article economy. It's still based on that. And Meanwhile, again, everybody else has experimented with a myriad of different business models and things like that and have moved forward in lots of interesting ways. And then to make matters even more kind of serious here, all of the reputation and credit models are based around the journal and the article still.
So I move this thing. The green thing. Yeah, I'm not a PC user, so I'm like, really dysfunctional. But the system of credit has revolved around journals and articles as well. Traditionally, and this is incredibly ingrained in everything that people do, but there are some sort of interesting shifts occurring with credit.
There are a lot of people now trying to recognize and reward all the outputs that people might be able to share. Thinking about other types of metrics around those outputs. Thinking about how early people are able to share. Some funders are doing interesting things to try to put policy around this or codify these things and reward them. And I'll do a shout out to Michele here in the audience for.
They've basically said to all of their investigators and their employees on the Janelia campus that they cannot include the name of the journal that they've published in. If they're presenting at a scientific conference, they can just put a Doi or a PMID. So they're just erasing the whole journal name altogether. They're saying that's not an actual proxy for the quality of the work. Let's remove that from and that distraction from the work itself.
It's bold. Aligning Science Across Parkinson's. One of the funders I've worked with is requiring the sharing of every single output with persistent identifiers in appropriate repositories and is counting all of those things and applying different types of credit to them. So the article is only one thing among all of the outputs that people are sharing.
So again, diminishing that importance. And then the Gates Foundation just announced they're not even going to pay apices anymore. So there's a lot happening where people are really questioning the value of all of these traditional credit modes. And this credit to this slide goes to Michele Whiting from HHMI. This is actually a technology called curve node that is at least evolving.
The article beyond the static article that we have with some links externally. This is where code data, things like that are embedded in the actual article. And this technology is already being used by, say, Argus notebooks. Now platform has come out. Norah Leber I don't think they're using curve. Note, but they're doing similar types of embedding Elis executable research articles based off of sensilla.
Another technology. I've been doing things kind of like this too, so we can break out of the mold that came from the truck. We can do more interesting things and provide a lot more value with all of those granular parts and pieces that Tim was blowing up in the last session. And then this is a silly image, but I kind of like it because it's a very simple way of understanding two individuals across the world from each other, doing real time collaboration around an in this case, a CAD drawing an engineering object.
But it represents and this is courtesy of a colleague, Jim colliander from to see. They build cloud based hubs for collaboration where people from anywhere can upload and use their data and tools all together in one place. This is a very breakthrough when it comes to science because it's changing like all of the different walls and silos that people had in the past to conduct their research. Now they can be adjacent to data and tooling from everyone else in their field.
So that's super interesting. So as the science shifts in how it's doing its work, the way that gets communicated naturally is going to need to shift too, because the very doing of the work is now generating the methodologies, it's generating the protocol, it's generating the data. In fact, people are reviewing each other's work in these clouds real time. That's much more exciting than peer review.
No offense, I used to be a publisher, but I think we should be embracing and thinking about the ways in which people are going to conduct their research is just fundamentally shifting in really interesting ways. And so with that in mind, how will all of this happening in the cloud across all of these different fields with people generating all of these rich data and code objects, commenting on each other's work, analyzing each other's work, at what point do we think we can still get them to open up a Word document and describe that.
I mean, I think it's going to be kind of hard to do that. So we really I would just say thinking outside of the boxes that we're currently in. And embracing some of these new modes. And with that, I will say Thank you. I'm not sure this ever happened before, but the speakers all stayed to the time they were supposed to. So that we would have 20 minutes for Q&A, and we do.
So any questions for any of the speakers. I hope so. Or we're done early. Yes, in the back. Thank you. Hold on a second. We're getting you a mic.
OK just. I better keep doing the bike shuffle. Yeah Thanks. My question was for Susan. Yeah I'm a my role is faculty writing coach at the Center for Faculty Excellence at the University of Oklahoma. So I advise faculty on writing related matters, publishing. And so on.
And I was very curious about the manifold interface with the forerunners and everything like that. Have you noticed any effect or have you heard any anecdotally even what effect that has on the writing process. Because I can imagine that can be quite motivating because instead of having this long marathon where you have something at the end, sort of interactive early going experience.
So Yeah. Do you have anything to talk address about that. Yeah so we've had well, first of all, I should clarify that we can do several different kinds of things from a timing perspective. In manifold, we can just produce the book of the print edition at the same time. Or we can do a project that is built over time. What we're finding is about 10% or less of the projects are faculty that want to build them publicly on the site.
So we might have 10 or 12 of those projects in progress. When we have 150 that are published works on the site. So it is a small subset of faculty that are interested in working in this way. The reasons we hear for people wanting to do this sometimes are that they want feedback as they're going. We have a senior scholar who produced a book and he just wanted to reveal the process of how he does his work. It was almost a teachable moment for him, but probably for other junior faculty coming up.
So what we're finding is projects do develop on a timeline of anywhere from 1 and 1/2 to three years. It's still quite a long timeline, but they're fairly accountable for their timing because you can see it. You saw that with Whitney's project, her draft project went up in April 2019, and that book published last year. So there was still a long time frame of building it.
She also had a research fellowship from the NIH that helped fund that time so that she was able to do that work. So I don't know that it really has changed the kind of credit. They get at the end. They still have one publication, but it certainly has made their work more public and readable as they go. And then they can also point to and use this when they're out speaking at conferences.
We have some scholars that use their manifold edition in their presentation because they can illustrate some of their points or offer people more when they're talking. So I'm not quite sure if that answers your question precisely, but that's what we're finding. Yeah sure. Let me see if I can show you the project. John let's see.
Here we go. Oops I got to figure out how to move this thing. Here you go. This is terrible. You've challenged my memory, and I can't remember. He's got three books going at once. It is not sensory features, although this is a lovely book.
John not. We have a lot of. Poor Whitney. It's every single reference with the name and John here. I'm going to have to get back to you on that one. Next question.
Yes Hello, I'm Joel Silver at delta think my question is actually a Christian, but I know her so I can give her a hard time. But it's actually. It's actually reversing to the others. May have ideas on this. I'm wondering if you can comment on your experience from the faculty side of how this is evolving, because I can imagine that if you look at academia and sort of where it's tied to the article, it's tied to the journal, the credit systems are designed around.
So are you seeing movement from academia and what they're doing. We'll see in the sciences because that's probably the bigger output and/or is there resistance and just they're still in the old world and everyone else is moving somewhere else. But I'm curious what you see. Yeah, I mean, it's a great question and it's really seminal because I think everyone who's trying to make change happen in this space, typically like your funders or even like heads of institutions see the culture change as the biggest challenge, and it's really entrenched.
And for all the reasons that we've been talking about, people are going to cling to that article. It's where their credit, their career advancement and everything else, their ability to get funded seems to lie. So the few funder decisions to do try to do things differently that I mentioned are drops in the bucket compared to the entrenched culture. But it's interesting like learning a little bit more about this.
So working with HHMI, for example, we did a survey of HHMI investigators and aliens and asked them and we did some workshops with them. And when they're asked, typically a lot of them will say that they see publishing as a necessary but annoying afterthought to the work. It's not really part of the process at all anymore. It's really this thing they have to do after they finished just to get that credit.
So they're very interested in other forms of credit, but I just don't think they're the ones that are going to change it on their own. They're just trying to do what they're trying to do and then that comes afterwards. But it is interesting, we also asked them what they thought of the new policies around data sharing that are coming out from the Nelson memo and everything.
And they're just like this is just like another thing. Now I have to publish and I have to prepare my data set and I have to share it. And it's hard and I don't even know how to do that. And where does it even go But then we said, well, OK, so new topic. What are you what are the barriers to collaboration for you. And they said, it's sharing my data.
And so what they meant though was they wanted to be sharing their data during the time that they're working, as they're producing it, when they have early findings with a more select set of colleagues. That's what they want to do. And that's where that cloud based kind of hubs and things like that come into play because then they get excited. They're like this is exactly what I wanted to be able to do.
This changes everything for me. So I'm really excited about that. And if we get that part right, then the rest of it can flow from there because the data already in the cloud, they're already in a space where you can just flip a switch and make them public. They're already in formats that are presumably, readable by others and things like that. So they're in this potentially fair space already.
So it is really just have to flip the way we're thinking about it. Instead of thinking how to hack from the article back, we should meet them where they are in their research and think about research communication as flowing from that. That's what I think. That's where I think this community here has an opportunity to really rethink.
Thank you. Next question. OK Jim has something to add there. Go ahead. Take that mic. So one of the components, I think, to the reluctance to move away from the PDF or from the article itself, I think we need to recognize this is that it's a form of communication.
And when there's communication, there's both a sender and a receiver and they have to go along hand in hand. The sender can't dramatically change the signal without making it very difficult for the receiver to understand what's being sent. And I think maybe this is why social media has been able to accelerate. So quickly because it doesn't really matter if you understand a picture of a cat or not, because it's very self-evident.
But research is very complicated. And so that has led to more inertia in the communication medium. And I think we need to. Include that with our thinking in that it's a sender receiver problem as well. It Hi, I'm Christine Fruin. I'm the director of publishing at the American Counseling Association.
And my question is predominantly for Susan, but we're happy to hear anyone else's feedback or ideas as well. So we are moving down a path of doing some experimental work in counseling research. And I have some members who really want to do more in interactive or digital publishing spaces. At the end of the day, however, I still have I have a revenue generating obligation.
And so we have been looking very, very seriously at manifold, either doing either a self install or a hosted version. And we've been looking at some other solutions as well. And the hurdle for me is, we're looking at we want to do some piece of it, open access, but some of it is going to have to be monetized. It's going to have to be. And just wondering if you've worked with any partners or anyone else in the room has addressed this on how to either monetize the entirety of the content or is there ways to make maybe the core text open, but perhaps folks have to pay some kind of access fee to maybe access updates or some of the innovative.
So just would like to hear either how others in the room maybe have experienced this or how you are. I mean, understand can bye bye. But what about buying access to the digital versions. Yeah so right now you could connect to a shopping cart and sell a digital version. You could do an excerpt, which we've done with some of our books on manifold. You can create reading groups that are private and then only offer access to people that would be given tokens into that reading group.
So that's another way that you could make that work. And we have had people do that. We at Minnesota haven't done it ourselves, because for us, this is our open access platform and that's what makes sense for us. But with that reading group functionality, that's one way to keep it in gated, if you will. I guess the challenge there would be scale. It depends on how many readers that you're expecting because manifold was intended initially to be open access.
We haven't thought through all the ways to gate it and haven't tried very hard. We've discussed possibly creating gated versions, of course. But it's not a direction we've chosen to go in heavily yet. And the book in question, by the way, is John Hartigan's social theory for non-humans. Social theory for non-humans.
OK hey, Adrian. Stanley this is for Kristen and perhaps Tim, but everybody else but the. The optimization of the article and blowing it up and changing and seeing research being done collaboratively and a lot of the work already done ongoing. What do you see the role of the publishers in the future. I thought I'd ask the question, but Yeah, I was trying not to get mean.
No, I mean, I think there's always opportunity. And so, as I said, if you stop thinking about continuing, I mean, I understand as a publisher and even though I was a publisher at plus and we were it was still all about the article and it was even more so, really. And there was very little room, very little margin for experimentation and really trying new things. And frequently when we did try new things like PLOS launched these publishing channels called PLOS currents, which were meant to be micro publications, and every author that used it wrote an awful article.
So like they had the opportunity to publish like data papers, cool, things like that. And they were like, that's not what I get credit for. So even when you try, you can easily fall into the same ruts. I mean, I think if I were a publisher today, I would work, try to work with the funders directly. I would just say, look, you're paying for this work. Let's open up a dialogue. Let's figure out how we can evolve in directions that make sense for what you are paying for and what you want to see happen.
Because I think the funder community is probably the most forward thinking and wants to reapply the way that they distribute funds in ways that innovate, that breed innovation. So I would take a step back and have that conversation and then figure out where publishing can change and how to become more relevant. And I do think funders are beginning to either become publishers a little bit or they're beginning to fund things that are alternative channels to traditional publishing.
And I would also keep a really close eye on that and watch those experiments because that's probably where things ultimately will head. Long timeline, no doubt. But if you wanted to get ahead of it, that's probably what I would do. Next question. Go ahead. Tim, go ahead and ask.
So I keep contradicting Kristen. Yeah, we have a diverse panel, a diverse panel. I work with Kristen all the time. So Well, I also see is a gigantic glut of AI produced content coming towards us. And I. We need high quality human curated content, human generated, human curated content to feed to AI in future.
And what may happen is that the literature, the scientific literature becomes increasingly contaminated with fabricated articles that can't be detected. And the publishers that are able to guarantee that their work is actually real and has been done by people and has been vetted. So this is a much higher level of peer review than we're talking about at the moment. That will become very, very valuable resource because it's vetted and not part of the great slew of very likely contaminated garbage, which you can't use to train your eyes because it's garbage in, garbage out.
And so down the road, I see I see this increasingly important role for whoever it is that takes the role of. Generating AI and human digestible information that is validated as being completely true and genuinely produced in the way we said it was produced rather than fake. Yeah all right, all right. Well, that's all very fine, but what if all that I generated articles really just diminished the value of the article because they start to look like junk.
And it was, Tim, who told me about some study where the public had more faith in science when data was also shared at the same time. So now imagine a future with all these AI generated articles. Everyone's like, this stuff is trash, it's worse than Twitter, whatever. And so now we want we actually want to go straight to the data. And what if the AI that applies itself to making sense and meaning making from all that shared data and code and all of that generates a much more valuable form of communication that then humans vet that absolutely replaces the article.
I hope I'm not making this up, but this is one interesting data point that occurs to me is I think they're on each other's board. He's not on. OK I was thinking it was. I was thinking it was a symmetrical relationship going on there.
And one other thing, but we do have a question coming up. But I'm just going to plant two words that I think we hear over and over and should pay attention to over and over. One is validation and one is credit. I think those two things are real drivers for what's going on here. Next question. Hi, I'm Shirley Decker.
Lucky I'm a director at SSR, which is a preprint server part of Elsevier, and this is actually kind of an interesting segue from the most recent conversation. I'm really intrigued by this idea of iterative output of scholarship and the way it can be changed and evolved in a way that it's shared with readers. I'm also very mindful of a lot of the conversations we've been having here about research integrity, and we currently, if we accept that our current default is one that's meant to be the most vetted, most peer reviewed, best we can do right now.
Authority of course, we know that doesn't always work, but that's our model. How does that change if we've got iterations, if it's a moving target, it's really hard for us to defend that against bad actors. But if we have a moving target of that research output, how does that affect our efforts to progress, research, integrity, defense and whichever panelist is interested in responding.
We talked a lot about this in the early days of making manifold and what we've defaulted to for now and this could change is that we note things as draft and it is not the version of record until we call it published. So as long as it's in draft mode, it is meant to be understood as a work in progress that could change and may be fixed, not just evolve.
And then that version of record then would be publication. And then, of course, there's the stage after publication where it could be added to and built on. Still, that's not the only answer, but that's the answer that we came up with for how we are recognizing things that are in development and not complete until the version of record comes along at publication. And I've got a comment to add on that as well that I think there's a real value in the fact that this new vision is highly collaborative by a group of participants that know each other and that provides some real security, a real ability to identify or reject bad actors because it will be harder to get in than this kind of more mechanical process we've got now where a paper mill can just.
Put stuff in and it gets through. Or somebody can fake an image, it'll get through if you've got. Pick a number 12 experts in the field that you're dealing with all working on this thing and something bogus comes in. They're much more likely to catch it before it even gets published. That's I'm an optimist. You can tell that, Tim.
I'm a pessimist. I worry about moving away from version of records. Because Well, for one for one reason. There's sort of a bit of a security blanket for researchers in that they can read the published version of the article, digest what it says, and half forget it or they can cite it in their articles. This article says this.
If you tell them you need to check back in a few months because that article may have changed and it may no longer say what it said. It says and then and so then they have this cognitive load that they're going to have to keep coming back to every single article they reference to make sure that it says they still say what they said. And then if it does change, they have to change their article, too.
And then there's a worry that the literature would be in a constant state of flux, of changes rushed through the literature over and over again. And it seems like an enormous intellectual burden is going to come up and disagree with me. Do we have time. Go ahead. All right. I will.
I will be brief. Far be it for me to shut up. Kristin written scholarship is always evolving. What we did with the trucks and the static articles and everything is we created an artificial view that there's this kind of thing that happens and now it's done. But that's not the way people actually do scholarship. And so we really need to reflect the fact that it's changing all the time and not make it feel like a penalty when somebody has to update their article there.
This is a solved problem. All the rest of the internet that we talked about, where versioning happens, threads happen. Everybody can see what the latest comment or discussion point was. You can trace it back in time. We now one thing we did really, really well in publishing is a persistent identifiers. They've solved this problem with version Doi.
So I think we need to be way more dynamic in our thinking. Well, we have reached. The top of the hour. So thank you very much.
Segment:0 .
Hi, folks. I'm Bill casdorph. I'm pleased to be moderating this session on beyond the article. I've done a number of presentations or written a number of things recently about the session originally called it is still called the diminishing importance of the article.
And I want to start out by emphasizing why. Why is this not advancing. Technical person still in the hallway there. A minute. Yeah that is. I got it. So, no, I'm not saying that the scholarly article is going to go away.
This is not about the death of the article. Article is indispensable to scholarly publishing. And just that it's been around for 350 years. Here's the first one you probably already know about that. Philosophical transactions 1665 still in publication. This is the world's longest continually published journal. And the article is just the gold standard for publishing scientists.
Scholarship has been free for decades and decades and still is to this day. This is the Watson and Crick paper announcing the announcement of discovery of DNA. My favorite spot in Cambridge. I did a lot of consulting with Cambridge University Press and I would always go to the pub where. He announced this discovery to his colleagues. And I sit-in the very chair that he was sitting in to do that.
I think that's pretty cool. But today, the article, I guess I would say is maybe necessary. I think Kristen may argue with that. She's the last speaker, but not sufficient because we need many, many more modes to communicate scholarship. This is really about scholars in general, but a lot of it is really centered on science. The key principle is the need to communicate about the research.
Throughout the research process, not just a year or two after the project is done and an article gets published. What are the issues. Well, first of all, as I was just saying, the article often isn't published until a year, maybe two years after the research is finished. And one thing that has been bugging me my entire career in scholarly publishing 40 some years is at negative results often don't get published.
And that's ridiculous because that's really important. Somebody else is going to duplicate that investigation. It's just important that got published, but it often is not published. The other thing that is commonplace nowadays is people realize that the research data needs to be available and findable. And we'll have Tim talking about that, or at least Tim's in the business of helping facilitate that, I'll put it that way.
But if you're really going to replicate the research, just the data isn't sufficient. You need code models, other resources. And ideally the whole concept here is that this sharing of this information and these resources needs to be done along the way as the research is underway and the text just isn't always the best way to convey some aspects of that. So just to reiterate, no, I'm not saying the article is going to be obsolete.
The formerly published peer reviewed article is still the gold standard in most disciplines. Notice I said, most not all. And progress in addressing the issues is being made. So here's just a quick run through of things that the speakers following me will get into some of these topics in more detail. But in terms of publishing the research.
Along the way and certainly not a year or two after it's done is obviously the growth of preprints and preprint servers. That's become very common. Some fields like math and physical sciences, they're used exclusively and oftentimes there is no actual final peer reviewed article. It's those practitioners, those researchers depend on preprints to communicate their research. And we've got a very good example of that.
I'll get to that in a second. And another thing that people sometimes kind of forget about is that there are disciplines where it's really conference papers that are the. Substitute in terms of validation and promotion in fields like engineering. More important than published journal articles, it's the papers that they live and die by in terms of negative results.
That's something that preprints can do. If some of you saw the session yesterday that's quite related to this new model from eLife is, I think, a really interesting thing where there's open peer review and the author gets an opportunity to revise. They don't have to revise it. They can just publish it that way. It lives, makes no accept or reject decision. The author determines whether they want to.
Respond to what the reviewers have said or not and whether they want to send it to a different journal or whether they want to put it up on eLife. And then. We all know that there are lots of data repositories. The difficulty there is in those three little words. Words? lots of two. Two words. Lots of data repositories, because it's kind of a mess right now in terms of where all this data is and how you can get to it and how you can find it.
So dataseer is an AI based system for helping determine what data to share, where to share it, how to share it. And that's Tim's brainchild. So he'll be talking to us in a few minutes. And in terms of models, Code Ocean has been around for quite a while. I think it was I think it was the winner of an ASP preview session about five years ago or something like that, I recall.
And that facilitates actually sharing the code and the computational research between teams. So it's a very useful thing. So my point here is these things are being done. This is not speculative or things like here's what we should be doing and aren't. It's like, no, we can and are doing these things just not as extensively as it should be done. There's a new format that I had the pleasure of co-chairing the Nasa working group that developed the CPLD standard.
It's a standard that's really designed for this because it enables packaging of arbitrary portions of content along with data and semantics and other resources into a single web based format that's optimized for interchange, search and display. And one thing that I think is really important about this CPLD format, it's quite new, so it's Elsevier uses it extensively, but it's not. It's just early days for this to roll out through the research ecosystem, but it's entirely based on web standards that everybody already uses.
So it's Schema.org, json-ld, some other W3C standards that may be less well known, like annotations and application manifest, et cetera. But it's entirely W3C standard based format. And then obviously, for different modes other than text for communicating research videos, 3D molecules, LiDAR, an archaeological dig, an interview with a famous poet, a documentary of a World War two battle.
Just to point out that this is not all only about science. So really the fundamental thing about scholarship is it's collaborative, and that's how we make progress. So I want to introduce I'm just going to briefly tell you who the speakers are going to be. And in what order, and I'll let them fill in any biographical details they would like to mention about themselves. But we're going to start with Stephanie orphan, who's the program director at archive.
And I think I'll leave it for Stephanie to tell you how long archive has been around so that you don't think this is a cool new thing. Well, it's a cool old thing is what it is after Stephanie to really demonstrate that this isn't just about science is Susan from the University of Minnesota Press. They've got a really interesting interactive platform called manifold that she'll be able to talk about and actually show, ideally, assuming we've got good connectivity that really enables that kind of dialogue and interactivity between people involved in a research endeavor, followed by Tim, who I've mentioned is founder and CEO of dataseer, which is this AI based service.
And I'm not sure that he's necessarily going to talk about dataseer. He's free to talk about whatever he wants to talk about. And he's shaking his head that, no, he's not presenting about dataseer, but you just need to know that that's who he is and. But apparently God was pleased about that.
And then finally, I will say the person that has really. Open my eyes to all of this for some time is Christina tan and she'll be bringing up the wrap up here. She's the founder of stratus, which is stratus strategies for open science. And she's deeply involved in this transition from. Static text based publication to dynamic ongoing publication.
Each of these presentations has got to be ideally quite short because we want to have plenty of time for Q&A at the end. So I will hand it over now to Stephanie. And I hope you know. You are not. You don't have slides.
I don't have slides. So bear with me on that. I have some very brief initial remarks and also very much listening to what my co-panelists have to say and how the conversation evolves after that. So I'm just going to talk a little bit about archive and preprints and I will start with that. So I think it's safe to say that archive was an early disruptor of the traditional scholarly publishing process.
For some background, back in 1989, Joanne Cohn, who was a physicist at the Institute for Advanced Study, had started an email list through which she was distributing scholarly articles in high energy physics or string theory, I believe, that were marked up in the tech format that really took off in the size of the list grew to such a size. I think it was like 180 people and she was individually handling requests.
It became untenable to continue on in that fashion. And a young researcher at Los Alamos National laboratory, Paul Ginsburg, offered to automate her email list. And that was essentially the birth of archive. So archive stayed at Los Alamos National lab for about 10 years until 2001. At that time, Paul moved over to Cornell University and he brought archive with it.
So for many years, archive was housed at the Cornell University Library. Many people still think it has a relationship with the Cornell University Library, but it does not. At about 2019, it moved over to the Computer and Information Science department and very quickly after that, by 2020, it migrated over to Cornell Tech, which is a graduate campus of Cornell University in New York City that's focused on primarily computer science.
And it does kind of like a marriage of business and computer science and things like that. It ended up there because the person that had been the Dean at kiss moved over to become the Dean at Cornell Tech, and he's a big champion of archive. So that's the history of archive and archive was and continues to be continues to be focused on the article is a research output.
But the introduction of preprints greatly changed the game. It allowed for fast dissemination of research so authors can get their research out there. Other people have access to it and can build on it. Very importantly, especially in the hard scientists facilitated claim staking. So a lot of people get that time stamp on archive and they can say, see, I thought at first I put it on arXiv, but over the last 30 plus years, things have changed.
Archive has grown significantly. We now have over 2.3 million articles. That's just like the number for like primary articles. In addition to that, there's probably double that in updates and versions and things of that nature. 60% of the content on archive is also published in traditional journals or through conference proceedings, but 40% of it isn't found anywhere other than on archive.
As I said, a lot of things have changed over the last 30 plus years since archive was born. I don't know if you've noticed, but preprints are having a little bit of a moment in the last couple of years. There's a growing number of preprint servers in a number of disciplines and some of them are discipline agnostic and there's a lot of experimentation currently going on around mechanisms for peer review as a basis, capturing comments using preprints as a basis for content for journals such as an overlay journals.
So that's kind of my little setup for what's going on with preprints and how we started this little shake up. But preprints aside, there's a lot of innovation and thinking going on in scholarly communications in general, and the other panelists are going to talk about that. And I'm interested in hearing what they're going to say. It's a really exciting time.
I have to stand on my tippy toes to see you all. I will do that. Sorry so I'm going to talk about the humanities and the social science perspective, especially from a University press. And there are others in the room that are doing similar things to me. I can see them out there basically at the University of Minnesota Press in about 2010, we were publishing a book called debates in the digital humanities, and we partnered with Matt Gould at the City University of New York for that.
And he wanted to do open peer review, and he wanted it to be a website that was accessible during the peer review process where everyone could read and comment and incorporate that work. And we created a website with him with an agency called cast iron coding. And the reason I'm telling you this story is because what we built for that project, that one book, eventually evolved into what is now manifold, manifold scholarship.
At the University Press, we had more and more scholars coming to us. Many of them digital humanities scholars, but not all of them. And they wanted to do things that were a little bit different than what they could do in a traditional print book. They wanted to publish in a way that was iterative. They wanted to publish as they were publishing or as they were creating the work. And I'm going to go into an example here.
Cut, copy, paste to show you an example of this. So what this scholar Whitney Turchin did with this book is that she started with resources. She had these collections of media and audio files, and she this is where her research began. So when we were building manifold, one of the things that we thought is what is involved in a book. And what we saw was that there were texts, there were media of all types.
And of course in the print book all you had was a static image. But in the web we have the opportunity to do lots of different things, including audio files data sets. And what she wanted to start with was an example of this. So this project for the first two years was just a collection of resources. And then the next thing that she was able to do, she was 41 total collections of resources here was that she started to write her book and we would work with her to clean up the text in a minimal way and put the draft chapters online.
And so you could follow the evolution of her research and her thinking as she was developing this book, because this was a new idea. She created this section on how to read this book for the reader, to orient them in what they should be expecting from her project as it was being developed. And eventually then of course, it became the full text that we know as the book.
And there is a print, there is a print piece that goes along with this. We have an actual book to go with this. So every manifold book we bring in an EPUB file and that becomes the base text. And then within that base text what we can do is annotate it with the resources and the media collection. So that's what you see in these side sidebar is these are where the media and the resources within her project fit within the text.
And then you can but we also have the original image from within the PDF and the EPUB file itself, and it's created a much more dynamic interactive experience. In reading the book. You can learn a little bit more about the different videos you can watch while you're reading if you want. You don't have to do that. We offer annotation and some different reader tools here and what we've done really, we think is given a more dynamic reading experience for.
Sorry It's hard to control and talk at the same time. What we've done is we can. We can create reading groups for classes. You, as a reader can annotate and highlight as you're reading, if you're logged into the manifold site. And we've created really an interactive reading experience in a different way than within the print book, classes you'll see have dialogues and interaction within the book itself.
So that there's some discussion outside of the scope of the class. And it has become a nice really. Tool for both for reading but for education. But then after we. Started with the book, then we decided to bring shorts series into it. So forerunners is a series of 35,000 words or less essays. Many of these might evolve into a full monograph, but this is another example of a semi formal publication that is an idea.
It's one idea that will eventually perhaps turn into other books and outcomes. But we allow our writers then a forum to have that discussion with their readers online. We've also brought in journals and. Which you see down here. Critical ethnic studies was our first journal that we put on the website. So in this way, what we do with these projects is we have a series of content blocks.
And so you can have texts, you can have syllabus, you can have media resources, and it's just a much more dynamic and multi-modal. Each of these things can be published as they are written rather than all at once as a final publication. So you can watch a work be developed and built as it's being published and written. And we've got about 30 different universities using it now. Cuny is using it for their whole system where they're publishing OERs where publishing papers where publishing class projects.
They have about 800 projects at this point on their manifold website. We also see Brown University using it for their campus wide readings. So they're bringing in a mix of texts and discussion and videos, panels with their scholars on campus and their entire campus. Then, as you can see with these annotations, are commenting and interacting as a community around publication.
It's one mode that we think addresses the need for the Humanities and social science, and we're excited to see what others are doing with their dynamic publishing. And of course, we'll continue to evolve manifold as we go. And now I'm going to turn it over to Tim. So we're coming back, back, back. Of course.
And just like the coyote. The Did we have a PowerPoint up there earlier.
OK, that's it. This one. Sweet and then we do this, I believe. But Thank you, Bill, for inviting me. So in contrast to the beautiful, well-developed ideas that have become products that have in some cases been in use for many years, this idea what I'm presenting is about as half baked as it is possible to be. This is an idea.
It's a concept. I think it sort of dovetails quite well with your CPLD. So why is this. So I've enlisted the help of Wile E Coyote. We're going to use Wile E Coyote to blow up the article and we're going to see what bits emerge and if any of those bits are of any use to anybody.
So we've blown up this Article I published by me in current biology, and here are the words I could have done letters, individual letters. But this is the bag of words, the first 150 or so of the 16 or 30,000 words in this article. And they're no use, really. I suppose you could plot a graph of how many there are of each, but it's not helpful.
If we use slightly less TNT, we might end up with sentences. Now, these are useful because. They tell us things. They contain subjects and objects, predicates and but there's no context. We're not sure whether these two sentences here. A related to each other and how they relate to each other. One is a sort of a statement of what we're going to do.
And then the next is the result of doing another thing. So if we take away even more TNT, we end up with blowing up the article intersections. And I'm going to argue that this point we actually get to something useful. And so this is a paragraph describing the application of a statistical test to a question and the result of that statistical test.
And we also find a graph in the article somewhere else in the smithereens. And we also find in the article a link to a data set and that data set is online on Dryad. And part of that data set corresponds to what we're talking about in the article text. So we have year, and whether or not the data were shared. And that corresponds to what's being tested here.
So now we're starting to put things together. And this reminds me very much of a semantic triple. That is we've put. We've got a sort of a question. We've done something to it. There's a process applied. And there is an object on the other side, which is not entirely unlike this.
So we've asked the question. To what extent is whether or not the data is extant predicted by the age of the paper. And we've brought data to bear on that question. We've ended up with a result. And this triple here, semantic triple is the raw material for machine learning. So I would argue that this three way combination of the section saying what's happening and what the result is.
And the result and the data that are used to test that. All the raw material that AI is going to use to understand science. We have a question posing, asking about the relationship between two variables. We test that and then we end up with an assessment of the strength of that relationship. And that can be Fed into an AI model along with millions of other triples to teach it stuff.
And so this way of structuring articles could become very powerful way to properly get AI to understand science. Because right now, to be honest, large language models are a bit like they've moved from asking a very intoxicated person at the bar what it thinks to ask your uncle that went to University. And so he'll tell you what everybody says about it the received wisdom on something but LLMs don't understand authority.
So they don't understand that or they give you the JPEG version of what the internet says about something. And so a lot of people are wrong on the internet. I don't know if you've noticed, but and so almost impossible to pick out. This is the authoritative statement on this and therefore this is correct. Whereas these triples. Give the strength of evidence underpinning an opinion.
And that's how LLMs can be taught anyway. OK done. Thanks you get. Good boy. Yeah, we are. Fuck You're ready to roll. No, that's Tim's.
That's Tim's bar last night. Get that. It's a little weird. You're not this beyond article. This one. Yep there you go. Hey, there's a lot of fret about this. Most of my fret about whether we'd get this.
Is it. OK Yeah, let me just. There you go. Awesome Thank you, Isaac. He's a miracle worker. Appreciate Isaac.
Hello, everybody. My name is Khristian rutten. And I as Bill mentioned, I'm the founder of strategies for open science, or stratus, which is a consultancy and implementation company in open science. And I work a lot with funders, some institutions, not so many publishers, but a partner with some. And then I also have a nonprofit called incentivizing collaborative open research, or I-corps for short.
And we started I-corps to do research on open science so that we can study ourselves and learn what works and what doesn't. So the evolution of media sorry is a really silly picture. I don't know why the guys are naked, but the guy with the computer at least should have clothes. It's totally weird. But I think we're sort of here when it comes to where we are with scholarly publishing. We're kind of at this laptop guy's spot where we're delivering things on the internet and maybe a phone or two.
But our formats are really still very, very print based. We are in the world of print legacy still 20 some years after putting journals online, we still are tied to the article, this container that made sense in a magazine or a book like presentation. We are still using issues like what's that about. The journal itself is like a book that had to get on a truck. So people had to gather up all the articles, create an issue, put this thing on a truck for it to go places.
So people could read it. We've lost the trucks are not involved anymore. So like but we're holding on to these formats and these kinds of structures and it's just not the way the rest of the internet has gone. I feel like there's been some sort of self-isolation. Let's see if I can move this thing in. Nope it says self-isolation and publishing partner. OK, so I think we've been in this kind of weird world in publishing where we started out along the same lines as all the other media with print and things like film and databases.
Things went online. That was like a big moment that happened. For those of us who are old enough to remember, that was before the web. And then the web happens and things started to bifurcate. And on the publishing side of things we embraced. And I say we because I was there participating in these things, committing these sins. So we embraced XML.
We thought HTML was too flaky for the gravitas of these things that were publishing. And it was the Wild days of the web and people were doing wacky things and changing web pages like from day to day or minute to minute. And that felt a little scary when you want this content to remain in its canonical form. So everybody got behind XML and the PDF made a lot of sense. We did things like we printed tables as images.
I remember doing this like it was pretty embarrassing. Finally, we got tables to be tabular data. We added some interactivity to figures zooming in and stuff and then links to all of these other rich resources that, as Tim was pointing out, are required to really understand the work, like data and code. Meanwhile, everybody else, literally everybody else went down completely different pathways and got all these goodies right.
So all of a sudden their apps, social media, all this information is happening across all these different platforms and formats. Augmented reality, virtual reality. And so I think we've lost a lot of opportunity there by having stuck with this pathway that really came from the trucks and has moved forward into this static state we're in now. And so we're stuck in this single mode and this has actually been really solidified through the business models and publishing.
And regardless of whether it's subscription or whatnot, it's still an article economy. It's still based on that. And Meanwhile, again, everybody else has experimented with a myriad of different business models and things like that and have moved forward in lots of interesting ways. And then to make matters even more kind of serious here, all of the reputation and credit models are based around the journal and the article still.
So I move this thing. The green thing. Yeah, I'm not a PC user, so I'm like, really dysfunctional. But the system of credit has revolved around journals and articles as well. Traditionally, and this is incredibly ingrained in everything that people do, but there are some sort of interesting shifts occurring with credit.
There are a lot of people now trying to recognize and reward all the outputs that people might be able to share. Thinking about other types of metrics around those outputs. Thinking about how early people are able to share. Some funders are doing interesting things to try to put policy around this or codify these things and reward them. And I'll do a shout out to Michele here in the audience for.
They've basically said to all of their investigators and their employees on the Janelia campus that they cannot include the name of the journal that they've published in. If they're presenting at a scientific conference, they can just put a Doi or a PMID. So they're just erasing the whole journal name altogether. They're saying that's not an actual proxy for the quality of the work. Let's remove that from and that distraction from the work itself.
It's bold. Aligning Science Across Parkinson's. One of the funders I've worked with is requiring the sharing of every single output with persistent identifiers in appropriate repositories and is counting all of those things and applying different types of credit to them. So the article is only one thing among all of the outputs that people are sharing.
So again, diminishing that importance. And then the Gates Foundation just announced they're not even going to pay apices anymore. So there's a lot happening where people are really questioning the value of all of these traditional credit modes. And this credit to this slide goes to Michele Whiting from HHMI. This is actually a technology called curve node that is at least evolving.
The article beyond the static article that we have with some links externally. This is where code data, things like that are embedded in the actual article. And this technology is already being used by, say, Argus notebooks. Now platform has come out. Norah Leber I don't think they're using curve. Note, but they're doing similar types of embedding Elis executable research articles based off of sensilla.
Another technology. I've been doing things kind of like this too, so we can break out of the mold that came from the truck. We can do more interesting things and provide a lot more value with all of those granular parts and pieces that Tim was blowing up in the last session. And then this is a silly image, but I kind of like it because it's a very simple way of understanding two individuals across the world from each other, doing real time collaboration around an in this case, a CAD drawing an engineering object.
But it represents and this is courtesy of a colleague, Jim colliander from to see. They build cloud based hubs for collaboration where people from anywhere can upload and use their data and tools all together in one place. This is a very breakthrough when it comes to science because it's changing like all of the different walls and silos that people had in the past to conduct their research. Now they can be adjacent to data and tooling from everyone else in their field.
So that's super interesting. So as the science shifts in how it's doing its work, the way that gets communicated naturally is going to need to shift too, because the very doing of the work is now generating the methodologies, it's generating the protocol, it's generating the data. In fact, people are reviewing each other's work in these clouds real time. That's much more exciting than peer review.
No offense, I used to be a publisher, but I think we should be embracing and thinking about the ways in which people are going to conduct their research is just fundamentally shifting in really interesting ways. And so with that in mind, how will all of this happening in the cloud across all of these different fields with people generating all of these rich data and code objects, commenting on each other's work, analyzing each other's work, at what point do we think we can still get them to open up a Word document and describe that.
I mean, I think it's going to be kind of hard to do that. So we really I would just say thinking outside of the boxes that we're currently in. And embracing some of these new modes. And with that, I will say Thank you. I'm not sure this ever happened before, but the speakers all stayed to the time they were supposed to. So that we would have 20 minutes for Q&A, and we do.
So any questions for any of the speakers. I hope so. Or we're done early. Yes, in the back. Thank you. Hold on a second. We're getting you a mic.
OK just. I better keep doing the bike shuffle. Yeah Thanks. My question was for Susan. Yeah I'm a my role is faculty writing coach at the Center for Faculty Excellence at the University of Oklahoma. So I advise faculty on writing related matters, publishing. And so on.
And I was very curious about the manifold interface with the forerunners and everything like that. Have you noticed any effect or have you heard any anecdotally even what effect that has on the writing process. Because I can imagine that can be quite motivating because instead of having this long marathon where you have something at the end, sort of interactive early going experience.
So Yeah. Do you have anything to talk address about that. Yeah so we've had well, first of all, I should clarify that we can do several different kinds of things from a timing perspective. In manifold, we can just produce the book of the print edition at the same time. Or we can do a project that is built over time. What we're finding is about 10% or less of the projects are faculty that want to build them publicly on the site.
So we might have 10 or 12 of those projects in progress. When we have 150 that are published works on the site. So it is a small subset of faculty that are interested in working in this way. The reasons we hear for people wanting to do this sometimes are that they want feedback as they're going. We have a senior scholar who produced a book and he just wanted to reveal the process of how he does his work. It was almost a teachable moment for him, but probably for other junior faculty coming up.
So what we're finding is projects do develop on a timeline of anywhere from 1 and 1/2 to three years. It's still quite a long timeline, but they're fairly accountable for their timing because you can see it. You saw that with Whitney's project, her draft project went up in April 2019, and that book published last year. So there was still a long time frame of building it.
She also had a research fellowship from the NIH that helped fund that time so that she was able to do that work. So I don't know that it really has changed the kind of credit. They get at the end. They still have one publication, but it certainly has made their work more public and readable as they go. And then they can also point to and use this when they're out speaking at conferences.
We have some scholars that use their manifold edition in their presentation because they can illustrate some of their points or offer people more when they're talking. So I'm not quite sure if that answers your question precisely, but that's what we're finding. Yeah sure. Let me see if I can show you the project. John let's see.
Here we go. Oops I got to figure out how to move this thing. Here you go. This is terrible. You've challenged my memory, and I can't remember. He's got three books going at once. It is not sensory features, although this is a lovely book.
John not. We have a lot of. Poor Whitney. It's every single reference with the name and John here. I'm going to have to get back to you on that one. Next question.
Yes Hello, I'm Joel Silver at delta think my question is actually a Christian, but I know her so I can give her a hard time. But it's actually. It's actually reversing to the others. May have ideas on this. I'm wondering if you can comment on your experience from the faculty side of how this is evolving, because I can imagine that if you look at academia and sort of where it's tied to the article, it's tied to the journal, the credit systems are designed around.
So are you seeing movement from academia and what they're doing. We'll see in the sciences because that's probably the bigger output and/or is there resistance and just they're still in the old world and everyone else is moving somewhere else. But I'm curious what you see. Yeah, I mean, it's a great question and it's really seminal because I think everyone who's trying to make change happen in this space, typically like your funders or even like heads of institutions see the culture change as the biggest challenge, and it's really entrenched.
And for all the reasons that we've been talking about, people are going to cling to that article. It's where their credit, their career advancement and everything else, their ability to get funded seems to lie. So the few funder decisions to do try to do things differently that I mentioned are drops in the bucket compared to the entrenched culture. But it's interesting like learning a little bit more about this.
So working with HHMI, for example, we did a survey of HHMI investigators and aliens and asked them and we did some workshops with them. And when they're asked, typically a lot of them will say that they see publishing as a necessary but annoying afterthought to the work. It's not really part of the process at all anymore. It's really this thing they have to do after they finished just to get that credit.
So they're very interested in other forms of credit, but I just don't think they're the ones that are going to change it on their own. They're just trying to do what they're trying to do and then that comes afterwards. But it is interesting, we also asked them what they thought of the new policies around data sharing that are coming out from the Nelson memo and everything.
And they're just like this is just like another thing. Now I have to publish and I have to prepare my data set and I have to share it. And it's hard and I don't even know how to do that. And where does it even go But then we said, well, OK, so new topic. What are you what are the barriers to collaboration for you. And they said, it's sharing my data.
And so what they meant though was they wanted to be sharing their data during the time that they're working, as they're producing it, when they have early findings with a more select set of colleagues. That's what they want to do. And that's where that cloud based kind of hubs and things like that come into play because then they get excited. They're like this is exactly what I wanted to be able to do.
This changes everything for me. So I'm really excited about that. And if we get that part right, then the rest of it can flow from there because the data already in the cloud, they're already in a space where you can just flip a switch and make them public. They're already in formats that are presumably, readable by others and things like that. So they're in this potentially fair space already.
So it is really just have to flip the way we're thinking about it. Instead of thinking how to hack from the article back, we should meet them where they are in their research and think about research communication as flowing from that. That's what I think. That's where I think this community here has an opportunity to really rethink.
Thank you. Next question. OK Jim has something to add there. Go ahead. Take that mic. So one of the components, I think, to the reluctance to move away from the PDF or from the article itself, I think we need to recognize this is that it's a form of communication.
And when there's communication, there's both a sender and a receiver and they have to go along hand in hand. The sender can't dramatically change the signal without making it very difficult for the receiver to understand what's being sent. And I think maybe this is why social media has been able to accelerate. So quickly because it doesn't really matter if you understand a picture of a cat or not, because it's very self-evident.
But research is very complicated. And so that has led to more inertia in the communication medium. And I think we need to. Include that with our thinking in that it's a sender receiver problem as well. It Hi, I'm Christine Fruin. I'm the director of publishing at the American Counseling Association.
And my question is predominantly for Susan, but we're happy to hear anyone else's feedback or ideas as well. So we are moving down a path of doing some experimental work in counseling research. And I have some members who really want to do more in interactive or digital publishing spaces. At the end of the day, however, I still have I have a revenue generating obligation.
And so we have been looking very, very seriously at manifold, either doing either a self install or a hosted version. And we've been looking at some other solutions as well. And the hurdle for me is, we're looking at we want to do some piece of it, open access, but some of it is going to have to be monetized. It's going to have to be. And just wondering if you've worked with any partners or anyone else in the room has addressed this on how to either monetize the entirety of the content or is there ways to make maybe the core text open, but perhaps folks have to pay some kind of access fee to maybe access updates or some of the innovative.
So just would like to hear either how others in the room maybe have experienced this or how you are. I mean, understand can bye bye. But what about buying access to the digital versions. Yeah so right now you could connect to a shopping cart and sell a digital version. You could do an excerpt, which we've done with some of our books on manifold. You can create reading groups that are private and then only offer access to people that would be given tokens into that reading group.
So that's another way that you could make that work. And we have had people do that. We at Minnesota haven't done it ourselves, because for us, this is our open access platform and that's what makes sense for us. But with that reading group functionality, that's one way to keep it in gated, if you will. I guess the challenge there would be scale. It depends on how many readers that you're expecting because manifold was intended initially to be open access.
We haven't thought through all the ways to gate it and haven't tried very hard. We've discussed possibly creating gated versions, of course. But it's not a direction we've chosen to go in heavily yet. And the book in question, by the way, is John Hartigan's social theory for non-humans. Social theory for non-humans.
OK hey, Adrian. Stanley this is for Kristen and perhaps Tim, but everybody else but the. The optimization of the article and blowing it up and changing and seeing research being done collaboratively and a lot of the work already done ongoing. What do you see the role of the publishers in the future. I thought I'd ask the question, but Yeah, I was trying not to get mean.
No, I mean, I think there's always opportunity. And so, as I said, if you stop thinking about continuing, I mean, I understand as a publisher and even though I was a publisher at plus and we were it was still all about the article and it was even more so, really. And there was very little room, very little margin for experimentation and really trying new things. And frequently when we did try new things like PLOS launched these publishing channels called PLOS currents, which were meant to be micro publications, and every author that used it wrote an awful article.
So like they had the opportunity to publish like data papers, cool, things like that. And they were like, that's not what I get credit for. So even when you try, you can easily fall into the same ruts. I mean, I think if I were a publisher today, I would work, try to work with the funders directly. I would just say, look, you're paying for this work. Let's open up a dialogue. Let's figure out how we can evolve in directions that make sense for what you are paying for and what you want to see happen.
Because I think the funder community is probably the most forward thinking and wants to reapply the way that they distribute funds in ways that innovate, that breed innovation. So I would take a step back and have that conversation and then figure out where publishing can change and how to become more relevant. And I do think funders are beginning to either become publishers a little bit or they're beginning to fund things that are alternative channels to traditional publishing.
And I would also keep a really close eye on that and watch those experiments because that's probably where things ultimately will head. Long timeline, no doubt. But if you wanted to get ahead of it, that's probably what I would do. Next question. Go ahead. Tim, go ahead and ask.
So I keep contradicting Kristen. Yeah, we have a diverse panel, a diverse panel. I work with Kristen all the time. So Well, I also see is a gigantic glut of AI produced content coming towards us. And I. We need high quality human curated content, human generated, human curated content to feed to AI in future.
And what may happen is that the literature, the scientific literature becomes increasingly contaminated with fabricated articles that can't be detected. And the publishers that are able to guarantee that their work is actually real and has been done by people and has been vetted. So this is a much higher level of peer review than we're talking about at the moment. That will become very, very valuable resource because it's vetted and not part of the great slew of very likely contaminated garbage, which you can't use to train your eyes because it's garbage in, garbage out.
And so down the road, I see I see this increasingly important role for whoever it is that takes the role of. Generating AI and human digestible information that is validated as being completely true and genuinely produced in the way we said it was produced rather than fake. Yeah all right, all right. Well, that's all very fine, but what if all that I generated articles really just diminished the value of the article because they start to look like junk.
And it was, Tim, who told me about some study where the public had more faith in science when data was also shared at the same time. So now imagine a future with all these AI generated articles. Everyone's like, this stuff is trash, it's worse than Twitter, whatever. And so now we want we actually want to go straight to the data. And what if the AI that applies itself to making sense and meaning making from all that shared data and code and all of that generates a much more valuable form of communication that then humans vet that absolutely replaces the article.
I hope I'm not making this up, but this is one interesting data point that occurs to me is I think they're on each other's board. He's not on. OK I was thinking it was. I was thinking it was a symmetrical relationship going on there.
And one other thing, but we do have a question coming up. But I'm just going to plant two words that I think we hear over and over and should pay attention to over and over. One is validation and one is credit. I think those two things are real drivers for what's going on here. Next question. Hi, I'm Shirley Decker.
Lucky I'm a director at SSR, which is a preprint server part of Elsevier, and this is actually kind of an interesting segue from the most recent conversation. I'm really intrigued by this idea of iterative output of scholarship and the way it can be changed and evolved in a way that it's shared with readers. I'm also very mindful of a lot of the conversations we've been having here about research integrity, and we currently, if we accept that our current default is one that's meant to be the most vetted, most peer reviewed, best we can do right now.
Authority of course, we know that doesn't always work, but that's our model. How does that change if we've got iterations, if it's a moving target, it's really hard for us to defend that against bad actors. But if we have a moving target of that research output, how does that affect our efforts to progress, research, integrity, defense and whichever panelist is interested in responding.
We talked a lot about this in the early days of making manifold and what we've defaulted to for now and this could change is that we note things as draft and it is not the version of record until we call it published. So as long as it's in draft mode, it is meant to be understood as a work in progress that could change and may be fixed, not just evolve.
And then that version of record then would be publication. And then, of course, there's the stage after publication where it could be added to and built on. Still, that's not the only answer, but that's the answer that we came up with for how we are recognizing things that are in development and not complete until the version of record comes along at publication. And I've got a comment to add on that as well that I think there's a real value in the fact that this new vision is highly collaborative by a group of participants that know each other and that provides some real security, a real ability to identify or reject bad actors because it will be harder to get in than this kind of more mechanical process we've got now where a paper mill can just.
Put stuff in and it gets through. Or somebody can fake an image, it'll get through if you've got. Pick a number 12 experts in the field that you're dealing with all working on this thing and something bogus comes in. They're much more likely to catch it before it even gets published. That's I'm an optimist. You can tell that, Tim.
I'm a pessimist. I worry about moving away from version of records. Because Well, for one for one reason. There's sort of a bit of a security blanket for researchers in that they can read the published version of the article, digest what it says, and half forget it or they can cite it in their articles. This article says this.
If you tell them you need to check back in a few months because that article may have changed and it may no longer say what it said. It says and then and so then they have this cognitive load that they're going to have to keep coming back to every single article they reference to make sure that it says they still say what they said. And then if it does change, they have to change their article, too.
And then there's a worry that the literature would be in a constant state of flux, of changes rushed through the literature over and over again. And it seems like an enormous intellectual burden is going to come up and disagree with me. Do we have time. Go ahead. All right. I will.
I will be brief. Far be it for me to shut up. Kristin written scholarship is always evolving. What we did with the trucks and the static articles and everything is we created an artificial view that there's this kind of thing that happens and now it's done. But that's not the way people actually do scholarship. And so we really need to reflect the fact that it's changing all the time and not make it feel like a penalty when somebody has to update their article there.
This is a solved problem. All the rest of the internet that we talked about, where versioning happens, threads happen. Everybody can see what the latest comment or discussion point was. You can trace it back in time. We now one thing we did really, really well in publishing is a persistent identifiers. They've solved this problem with version Doi.
So I think we need to be way more dynamic in our thinking. Well, we have reached. The top of the hour. So thank you very much.
Form Title Download
Transcript
Form Title Bookmark
Entire Media
Segment(s)
Custom Clip
Start At:
End At:
Form Title Share
Entire Media
Segment(s)
Custom Clip
Start At:
End At:
Form Title Accessibility and Keyboard Shortcuts
The cadmore media player is a professional video and audio player experience designed to deliver media as well as relevant transcript, segmentation, and metadata. On the screen is a traditional html video player with buttons such as play, pause, forward, captioning, language and play speed selection and more. Adjacent to the video player are generally tabs, one of which shows a scrolling transcript of the video and the other shows any segmentation of the media, sometimes referred to as chapters. At the top of the video is a menu bar containing an Explore button and Tools button. The Explore button allows you to dig deeper into the video by opening up an explore dialog which contains tabs for a visual navigation of the video using thumbnails, an about tab which contains rich metadata about the media, a segments tab which contains further information about the video, and a related tab which contains links to other media related to this media. The tools button opens a menu with options for sharing the media on social media, creating bookmarking links, creating embed codes for your website, generating citations, and more.
Keyboard Shortcuts
Keyboard shortcuts only work while you are in forms mode. By entering forms mode, you will be able to quickly access portions of the media player with single key commands.
Spacebar: Play/Pause Toggle
I: Info Menu,
O: Tools Menu
T: Transcript Tab,
S: Segments Tab
F: Forward 15,
R: Rewind 15
L: Languages,
P: Playspeed
M: Mute Toggle,
V: Volume Bar
N: Video Only Toggle,
B: Full Screen Toggle
C: Captions Toggle
Escape: Closes dialogs and menus / Exits Player
Z: Next Transcript Chapter
X: Previous Transcript Chapter
Q: Next Transcript Paragraph
W: Previous Transcript Paragraph
Entire Media
Segment(s)
Custom Clip
Start At:
End At:
Embed Size
-
You Size - Responsive
-
We Size - Responsive
-
Picture Overlay Popup
-
400 x 225
-
512 x 288
-
400 x 700
-
768 x 1024
-
1024 x 512
-
1280 x 608
Form Title Cite
No citation provided.
Info
Segments
Related
Thumbnails
Exit Clip Mode
You have requested to see a portion of the video outside of the clip.
Click OK to switch to the full video or click if you would like to stay inside the clip. You can always go back to the clip with Tools and Clip Mode