 So, I don't know where to start, actually. I thought my role was just to sit here and look pretty, but I guess I have to ask some questions, too. So, I guess before we start, so I'm actually a researcher and I try to build also some kind of data infrastructure for computational chemistry. And it's a difficult problem. So infrastructure is not a sexy problem and actually nobody really cares about it. So, the question is like how to get money and that's also how I ended up here and looking at decentralized solutions. So, I think I would like to say that I'm not also particularly how to put it. I think decentralization is probably more important than blockchain because science is fundamentally decentralized activity. So we need to, I think, develop infrastructure which supports this kind of effort. So it doesn't matter if you are from Brazil or Max Planck, you should have the, you should have ability to access data and from everyone in the, from anyone in the world, right? So how do we do this? And I guess what are your thoughts about that? Who wants to start, Dennis? Hello, hello, yeah. My interest in preservation of scientific data came from interest in open data. So like in Ireland, we started to, like when we introduced the Irish government to the concept of open data, there were exactly zero open data sets. And then, like probably in 2015, when there were a few thousand data sets, we asked the question, how do we know that, first, this data needs to be trusted and secondly, how do we know that it will be available tomorrow? So, and we started to think about how to employ those distributed technologies for preservation of this data. And coming back to my road analogy, like I think it needs to be, the infrastructure has to be publicly funded. I don't see, like because the projects have the beginning and the end and if you go from one fund or like one grant to another, like as soon as you get one grant, you start thinking how you get another one to keep it going and that's not sustainable. So, like building anything starts from standards in our view and yeah, defining first of all what needs to be preserved and what is valuable in the future, that would be a good start. Yeah, how to do it, it's an open question because there is communities in Silas talking like one community will discuss research object, others will work on protocols and yeah, experiment with something. I don't think that those blockchain communities have enough librarians in place and scientists and practitioners. And at the same time, yes, there is not enough people from a technological background go and talk to the people who discuss and research objects and protocols. Also, I wanted to say if anyone has questions, please just also feel free to ask because I think we can make it more interactive like that. If you have any suggestions or if you're thinking about anything. Otherwise I can continue, yeah. And in a way it's funny because we just had the discussion at dinner so definitely here are also some of your ideas that we discussed there and yes, infrastructure is unfortunately not sexy especially as you said, actually establishing this is even easier but then the maintenance, this is actually the tricky part that is hard to get money for and luckily I think there are new things on the horizon that I think more long term but also standards are critical. And why it's not sexy is because it's effort and as a scientist, I'm a scientist myself and working in an infrastructure institution now, I see the issue or maybe even the solution there that we have to make this infrastructure thinking sexy again by providing cool tools. Currently people cut out images, put them into PowerPoint and store this as their data and this is ridiculous. This is just generating breaks in the scientific workflow and if we can convince people that these cool tools help them to make their research better, easier, faster and more reliable then we might gain their interest and maybe also their money and then they or not their money but that they tell the funders and by the way I need more of these cool tools in order to make my research better. So in the meantime, do we start first making these tools or actually mobilizing people to actually ask funders to support making all these tools? Yeah, it's a chicken egg problem clearly. My approach would be start with cool tools that you build from stuff that is around and sometimes you just have to put some stuff together just to show, okay, if we do it like this, you save already some time. I hope so. Any ideas from, yeah. Yeah, I mean, there's a lot of cool tools already that exist like think about the Kaggle community and all the bounties that are open there and the challenges and basically working on honor and rewards functions. Okay, I guess then let's reformulate the question. How do we make researchers use those cool tools? Well, I mean popularizing these things in university shouldn't be too difficult to give them a bit more time or make them, like what if research bodies or funding institutions create such kind of challenges for people to solve. But that's kind of the problem because it is difficult because nothing really changes or it changes very, very slowly. So some cool tools have existed for many years and I think I gave an example of Google Docs. So Google Docs exists for many, many years and still I personally got like version 12 of Wordfile and version 12 and then initials who last changed it so we don't even use Google Docs to actually write papers. So I think cool tools are developed in the outside world but we are kind of stuck like close to the, I don't know, paging. I can't remember, like I did five years of research and indeed I got those version 12a from Wordfile in email and then I had to change all the styles and it was a mess. And then you open it on your Windows machine and it was a Mac file and then all the formatting. Researchers don't have Windows. I did discover Bitcoin when I was researching and other type of cool tools or emerging things. So yeah, I guess that's more like inside of the researcher itself, like his ability to become more, he or she to become more productive or use tools that are enterprise grade or, I don't know. Yeah, how to change the mindset of a researcher basically. I don't know. It's also a legacy problem I think because yeah. Yeah, I'm sorry, I'm gonna be the one that always brings us up here. I'm just, let's play a thought experiment. You guys are all researchers. You get into your office the first day, you need to do a bunch of experiments. You're probably gonna spend 14 hours on those experiments. What are the actual problems that you're having right now to your life for you to feed your family? So it's funding, right? What do you need to get funding? You need to get published. If we go back to first principles about all the things that you do every day as a researcher, you all were researchers at some point, don't forget that, right? Then you find the solutions, then you know what the actual problems are and then you can actually find the solutions, right? Getting a researcher to change their mind, it's not hard, but what's the point if I'm going, all I wanna do is look at collagen forever. That's the only molecule that means the most. Why do I care about blockchain? All I wanna do is do my research. Well, that's exactly the thing. So you shouldn't care about all these things at all, but you have to because at the moment there are no other tools than the existing ones, right? So that's why you have to think constantly about funding and publishing because that's how the system is designed. So we need to change the system, but the question is how and we need to change. Exactly like if tomorrow, let's say, we don't have to build anything new, we just need to make people nicer, essentially. So that you don't, it is, it goes back to actually making people nicer. So you don't, I don't know, publish crappy papers, for example, right? So you don't have to publish as much, you don't have to maybe chase so many grants if grants will last longer than a year or two, right? So it's also about how do we support research? And the truth is we don't. So I think this conference is about actually finding possible solutions, like how can we support research? And not in a way like what can we do to fix the current system, but my thinking is like what can we do using this new technology to actually make something new and different and actually more aligned with you, I mean personal values of the researchers, which is mostly like curiosity, and like that, yeah, there are people who are interested in collagen, for example, and why not let them to actually spend their lives researching collagen, but you can't do that. Doc, you have to research that. And then a few other things that are, that you know that will get funded. Yeah. Well, just to tap into this a bit, what I figured is that these decentralized communities is all about the communities themselves. And what blockchain brings is not only just a community of messaging, it's also a community of sharing actual value and redistributing value and getting provenance in place and all these things. So the actual tool is community building. And if you're a researcher that doesn't care about what other researchers are doing, just go for your next publication, then you're probably gonna have a hard time anyway. So if you can live in a collagen community, where you share insights and issue mutual rewards or whatever you have, I think that's basically the start, no? Community thinking. Just as an idea is a solution, researchers work really well if they have a mentor. And that's not a new concept, but maybe the mentor needs to be the person that has that link to blockchain. So the researcher doesn't have to get involved in that aspect. There's somebody who takes that away and does something with it. So you have a bridge. Well, that's a really interesting comment. So something that I, when I look at all the solutions in blockchain space for science, there's a lot of incentivizing of peer review, which I think it's a futile effort because there's no really need to incentivize peer review. People will gladly criticize you. But mentorship, for example, is very important. And nobody's looking actually to create incentives for people to be good mentors. And it's something that will first, well, essentially because researchers are so overwhelmed and something's gotta give. And this is like one of the first things that are going to crack because you are not really rewarded to be a good mentor. So a lot of researchers actually, or young scientists are not really being mentored very well. So that's also one of the things to look at. But in terms of, I guess, data, I guess, that's again, a different problem. So it's all kind of connected, but how to solve everything in one go, I don't think it's possible. So we need to try and experiment. I think this is all one big experiment. So we don't also have to be overly critical because it's very young technology and many things will fail, I guess, as well, but maybe something good will come out of it as well. To have a chance, we can create many technologies, but they have to be used. And for them to be used, it has to be like the solution have to fit into researchers' workflow. And you shouldn't think about what blockchain is and how it's saved. It's not your problem. You want to build career and to feed your family and to pay you to buy a house, whatever you want to. And it should be possible to do on data attribution. And like right now, there is increasing demands whether from funders or from like from European Commission, for example, if you get more than a certain amount of grants like of resources for research, like half percent of it needs to go for data preservation and each and every proposal has to have data management plans. So that's one incentive. How the solution and how the tools will be built and how obviously there is like different disciplines and different tools required, but it's all needs to be built on, like all those tools needs to be built on protocols and on data. They have to be interoperable, right? Yeah, interoperable, yes, because if we want to have a chance to employ artificial intelligence, data landing, or I call just algorithms, they're all driven by data and if you put rubbish in, you get rubbish out. So maybe James can tell us a little bit more about this linking data and creating knowledge graphs. So there are a couple of speakers who comments on the floor. It was like we're all researchers, but we're all different types. So I'm in a different camp, I think, for most of us, I'm a citizen science, DIY, do it yourself. So we have a completely different view and it sort of calls us like the equivalent to the unbanked in Bitcoin. So I think we will go for this self-sovereign identity sandboxing approach. So this in this extreme form, we just eliminates the data infrastructure problem because we don't share any data, we keep it to ourselves, but it puts all the onus on the computation, as Dimitri said, and that's still in bionic form, but I think WebAssembly will come and give us that security. In terms of, even as a citizen science, I'm doing it with wearables, the use case that Konrad mentioned, and I've got various types of brands of devices, but they're all heart rates, right? So I have to unify the heart rate data type. And so tomorrow I think I'll take one of these 10-minute slots and talk about a coded project, a pseudocoded talk-through project called HeartChain that guarantees cryptographically across the community a data type, and a data type that then encrypted to the raw data coming off a device. And the third thing I'd mention is that there is a lot of money and data, you only have to look at the ICO of Filecoin. And those people do reach out and come to our events to be fair, but I would say we're getting a very tokenized and contribution from them because I've got some heavy engineering to do. My rebut back to them is that if you may deliver a solution, but if you do it in isolation from the science community, it may or may not be the right fit for us. And I guess I also have one more question probably for Dimitri and Konrad, and I think it's also a very important one, and that's data ownership. So it's very complicated. So universities make claims on your research data. You have certain rights to your data, then there are national legal, there's national legal framework, funders make some claims, so who owns it in the end? And what are your thoughts on the problem? Well, that's definitely a tough one. I mean, what is missing also the patient, right? If you have kind of a clinical study of the patient in there, in the end it's his or her data and he or she might donate this, but maybe to a certain restriction for that piece of research, but maybe then another one, another research endeavor pops up that would like to reuse this. That's a super tricky issue and it's also between different jurisdiction and different different countries. This is super tricky. I even think we've, once we have the technology solved, this will be one of the biggest issue, harmonizing this in the different countries. I mean, with a European Union, you might have one chance to get this done in a right way, but I think we are far, far away from solving this unfortunately. Yes, I have two comments. One is regarding the blockchain solutions which are now designed for helping science. We hear different stories and they sound like they are the ultimate solution. What if we have a solution which is based on different blockchains or not blockchain-based solutions, but they are interoperable? We did not hear so much so far about this, that there are multiple ways to solve this problem and what if they could intervene with each other? They can, yeah, we can choose one solution that's completely fine and we can retrieve data with the other method. Is there a possibility for that? I think that's probably the solution we should aim for, right? As was mentioned before, like that every research field has different needs and we don't need the same tools, right? So we should have something that we can actually customize to our own needs, but we have to make, I think we should allow that these different platforms are able to communicate, so that they are interoperable. So for example that chemistry researchers can communicate, for example with medicine researchers, right? About drugs and patients data so that can all, but we don't necessarily have to use the same platforms because obviously we'll have different requirements because chemistry research has different requirements compared to patient privacy and clinical trials, for example, but it is connected and I think that's how we need to make it flexible and interoperable and I think that's the key and now whether it's based on blockchain or something else, I think it's less relevant as long as it's possible to exchange the information. Yes, I agree with that and the second comment is that we heard the technological part of it and we hear that we want scientists to go along and use these fancy tools. Why don't we take scientists into the development? Why don't they choose what they want to do with it? That is a very good comment and as a researcher trying to build infrastructure, that's what I came to realize because oftentimes you have funders who kind of hire someone to develop certain tools which are then completely useless because they are not really developed with researchers, for researchers, they are just, here you go, use it and it doesn't really fit into the people's workflows. So that's why I said it needs to be flexible so people can customize it but it also needs to be easy to customize so you don't have to go and learn everything about the blockchain, it should be possible to just do it in a very, or even have dedicated people but then we go back to the problem of maintenance and support for the infrastructure which is just, it hasn't been the case up until now. So we have a unique opportunity to actually build something new and different and think about what it is that we actually want, not try to copy the existing solutions and put them on blockchain which I think would be a very bad idea. May I add to those points? I think this was done in the previous time that the solutions were put out there and now the scientists just have to enter and play around and they didn't come so this is the issue. So I think meanwhile also the funders see that there has to be a community-driven approach to that so it's not that you just build infrastructure by yourself but always a community needs to be embedded when doing these decisions. And regarding your first question, it's, well, having just one single solution would be great but I doubt that this will work out. This is, we need to have kind of an agile approach here because we don't know where we end up so we need to have different solutions and then kind of a selection of that and consolidation of stuff that worked out and that didn't and keeping the chance to communicate between these different solutions would be then essential for that but we need a kind of competition for the best solutions and it's still kind of research, right? It's science, right? We don't know currently where we want to end up and it's an open question and this is why we have to play around in different sandboxes. And I guess I have one more question for Dimitri regarding the ocean protocols which I think is really exciting and an interesting idea but how do you plan to actually bring people do you have any, what are you doing to actually engage people into using it or spread the word? Yeah, so, well, we have a lot of bounties and ambassadorships and all these things and basically saying that if you contribute to the network you become a stakeholder in the work, you get equity, early mover rewards there is a lot of basically gamification around that as well and also like doing all these speaking, conditioning. Is it working? It's working, it's perfect. We have, like we issued in the last two weeks we issued three bounties and they all got already like within 48 hours they got fully accomplished and we issue our native token so I think for us it's like an initial experiment but very promising. I want to mention about interoperability and other things like the reason why I think it's very important that you have to look at what's the trade-off in each situation do you want consistency, partition tolerance availability of resources then you have to choose different type of tools but I think the clue to all these things is a form of abstraction like you would see for example how the internet is built there you have like the OZ model you have like seven layers of abstraction such that it becomes a very flexible system where every configuration is almost possible and you don't have to care about the underlying network just abstract it in such a way that all these components are seamlessly working you don't have to care about how your model works in order to send a packet over it in order to create a connection in order to send an application due to an HTTP request it's all different responsibilities and I think that's really how the decentralized stack is working out creating those levels of abstractions what we are missing is an end game creating that nice user experience because that's where the bigger companies win that's where the Google and the Facebook are stellar they create the user experience that's less worse than what you see for example I use GIMP instead of Photoshop honestly I'd rather work with Photoshop but yeah you get GIMP and there's a lot of these things that are in open source and from researchers but they just miss the final edge and that's the UX and that's basically also where most of the bigger companies get their advertisements in between or what have you the user behavior steering function so if you find a solution for that maybe we also have a last mile of open source towards the end user I think we are running out of time but yeah I think definitely we need to invest more in the user experience which is for some reason completely neglected in almost every research application thank you for your attention and if you want to talk to any of us just grab us in the coffee break