 And now I'm very happy to introduce the next speaker, the first speaker. It's Professor Diernagel from the Department of Experimental Neurology of the Charité University Hospital. And he is known in the open science scene very well. And I just hand over the microphone to you. Thank you very much. So I hope you won't regret having me invited, especially as the first speaker, because it will be controversial. And so if you fire this up, my slides are on OSF, as always. So there is a URL. So my slides are there. I should start with a disclaimer. And the disclaimer is that I'm not going to talk about blockchain and science in general. I think they'll provide a lot of useful and interesting applications that are coming up. What I will talk about is the context of the blockchain in terms of increasing value in science. And I took this from, this is the abstract from the blockchain for science and knowledge creation living box. So it must contain the most important things. And actually, it only contains two aims of the blockchain for science. One is to, I don't know, cause a sociocultural, legal, political, economic revolution. I won't talk about that. And the second aim is that it would make larger parts of the research cycle open to scientific self-correction and that it has the potential to be a new approach to the current reproducibility crisis and so forth. So this is a pretty strong statement, I should say. So this is what I'm going to talk about because I'm very skeptical about it, I have to say. Interestingly, and I'm very happy that this happened a week ago, Kevin McKernan at the Texas Bitcoin Conference had a talk. Not sure, maybe even some of you saw it. And the interesting thing about this talk is that it filled this promise with real life because he was actually presenting a project, a fully crypto science project, which he thought would substantiate the claim. And the claim he made is that crypto solves all those problems and all those problems who enter reproducibility, blah, blah, blah. And he even said that the distributed ledgers are a requirement for the scientific method, which is, I'm glad that Newton, Galileo, and Einstein, and Watson and Crick didn't know about this because we'd be not where we are now. So I'll come back to this because he actually made the same claim. And he thought that he's going to tell us how he did it. And so I'll return to this. But before we get there, just a short reminder. And unfortunately, I cannot go into this because this is a blockchain conference and not one about the problems that we have in science right now. So I'll just remind you of a few of the problems that at least a few people think we have in biomedicine. And I'll breeze through it. So in the Lancet, there was a series that basically said that 85% of the health research is wasted. I was part of that series of articles. There is a general notion, one should say, about many scientists that we have a problem in reproducing of what we are doing, which I think is real. And this guy, John Ioannidis, in 2005 basically said that most published research findings are false. And he didn't just say it, he demonstrated it. It's a mathematical paper. It was prescient in a way because he said what he basically says is it's about bias and it's about low statistical power. And he was predicting that less than 50% of what you read is true. And that I think is pretty much substantiated by what we now see in terms of reproducibility. Now, I really have to breeze through this, but it's important that I tell you what I think are the problems we're having. And I have to admit that this is a very biased selection. You could come up with another one or you could even say there is no problem. But this is a selection of things which I think are important. And the reason why I'm reading it to you or telling you about it is because these are the things we need to solve if we want to solve the problems. So there's tons of bias, selection, detection, attrition bias. I'm talking about biomedical research. So this is we're trying to find new drugs. We're trying to find mechanisms of disease and so forth. So there's lots of bias in this type of research. There's research of flexibility. So we use data very selectively. We switch outcomes. We heart is hypothesizing after results are known. We do a lot of statistics to find the P values that we want. We have insufficient statistical power, meaning there are not enough ends in those experiments, be it animals or be it humans. We confound exploration with confirmations so most people do an exploration and then they say they confirmed something. They showed a mechanism. Actually they came up with a hypothesis but they didn't prove it. We have a lot of methodological problems that are not going into them. There's a lot of storytelling. If you read a paper in cell or in nature, most paragraphs start with next week and they give you a narrative. Like they went through it from sitting in a bathtub or under the shower having an idea. They went to the lab. They did the first experiment and in the end they had this beautiful paper but it's actually just painting on a canvas with lots of things that they did quite often very unrelated and selectively used. We have a file drawer problem which is substantial so stuff is not published. People don't, if they don't want, if the results were not what they wanted, they don't publish it. And we have low external validity meaning a lot of the models that we use do not reflect the diseases that we really are interested in. So this is, I could talk about each line here for two hours but you would not want me to do that. But why do scientists do this? Are scientists stupid? No, I mean these are professionals. The reason why they do this is because, and that's the cause of the cause in my view, is, and that's a root problem you could say, we have the wrong incentives and we have misconceptions in academia. These are the root causes. And the rest is just the result of it, what I've shown you. And that's another list which are not root causes but these are sort of close to the root causes. We have a published in Parish Culture and these are just reminders. I think most of you know what I'm talking about. We have perverse metrics, most of them are proxies like impact factor and stuff like that. We value competition over cooperation. We have a dysfunctional peer review. We have a mantra for excellence which is completely silly. We have a Matthew effect so if you have, like I have an institute and so forth, you get more. And we have a total misconception about whether innovation can be planned or not. So we have lots of programs where we should come up with something new but this is not how it works. But this is a package and there are many more things and it's kind of, it's part of the root problem. Now, what are suggested solutions and these are just technical ones. This is also a list that's my list if you wish. I share this list with a few people but nevertheless you could come up with other things. And again, these are the technical things like you have to educate people. You have to help researchers with all kinds of things. You have to increase sample size. It's kind of a reflection of the previous list. You have to replicate. You have to publish null results. You have to be transparent about what you do. You have to pre-register your studies, et cetera. And I think we need to do meta research meaning we have to look at what we are doing to find opportunities to improve it and also to show that if, and what we are doing is actually making things better or whether it's making worse. But all this stuff, this technical stuff won't help if we are not tackling the root cause. Otherwise it would just futile. And so the root cause is, as I said, I think it's about incentives. So we need to change the incentives. There are easy fixes, fairly easy fixes like awards, badges, stuff like that. And it's not very effective. It's easy, but it's very effective. It's not very effective. And there are the hard ones, but they are the key ones. So we need novel criteria and responsible metrics in assessing ground applications, researcher careers and so forth. And I mean, you can come up with, I would say crazy stuff. If people do it, some of it is not that crazy. We need to disseminate our results to our institutional repositories. We need open post-publication review. We need, I think, more base funding for researchers so that they can play. And we need peer-to-peer funding, maybe lottery, stuff like that. So there are lots of things. Now, so this is my list of how to increase value. Now, let's get back to the blockchain. So remember, the promise was that the blockchain technology will improve all this. Now I'm coming back and I'm very glad that McKernan had this talk, because I think he did a very good job. In fact, it was a brilliant talk about plant genomics. The project he presented, and I really urge you to watch it, because it's nice. He explains plant genomics. He did a cannabis sequencing special cannabis plant. It was kind of a disruptive project because he wanted to kill some patents so that people could propagate this plant for therapeutic uses. Nice project. And he also explains the science in a very nice way. But the reason why he did it at this Bitcoin conference is because he claimed that this is a blockchain project and that this is open source and that this is crypto. So that's what, these are the base claims. He said the funding went through Dash Labs. And many of you know that. It's a data fidelity thing, which gave him also the opportunity to publish negative results. The peer review was open and crypto incentivized, and the preprint also was crypto. Great. So this is the claim, the whole scientific process in the blockchain. Now let's look at it, what he really did and what he could do. First of all, the funding was, this is a cryptocurrency company. I think this is crazy. I mean, it's nice for him that he got the money, but no cryptocurrency company will fund my stroke research. And I'll come back to that. It was paid out in dollars. Now the data fidelity, the negative results stuff, obviously you can't publish the true data in the blockchain because this is just too much stuff. So you put in hash, you know where he did the data, where he put the data on the IPFS, mega, NCBI, AWS, and so forth. This is where the data was. So I don't need a blockchain for this. The peer review was, he said, open and crypto incentivized, he gave $1,000 for a review. Now this is highly, highly up to the discussion, whether we should pay for a review. And I actually have to say I'm against it. But I'm for open review, but this is also highly controversial. But I don't need a blockchain for this. I mean, in fact, the review didn't happen on the blockchain. So just the incentivation happened there. And then he was talking about a preprint. It turned out in the discussion that the preprint was on the open science framework. Sure, it had to be because you can't read a blockchain. So now if I would do the same project, what would I do? First of all, I would try to get money from public funding because the public has to fund our research. Well, of course companies can also fund research, but I'm a taxpayer and I want to have my taxpayer money to go into useful stuff. So I rather change the NIH or the DFG funding. Why would you start something in parallel? There is a funding stream and we should get part of it for good stuff. The data fidelity. Well, as I told you, he put it up on IPFS and Mega and so forth, but there's lots of more. There's the OSF, there's Synodo, there's the European Science Cloud and so forth. And they're all time stand and they are immutable and so forth. Why do I need to put it on a blockchain? Peer review. Well, the peer review. I mean, it's already there. There's F1000 research, there's Royal Society Open and so forth. It's not monetized, but I'm against monetizing this process. I think it's wrong. And the preprint, I already told you, he did it on OSF, he could have done it on BioArchive. So what about the blockchain in this context? Again, I'm not, you can't be against the blockchain and there's useful stuff, but it's not about value. So beyond phrases, it's a technical tool and we need a culture change in the research that we do. Our research, that's also important. I mean, this is a bubble, you're sitting in a bubble. Our researchers use paper notebooks. They scribble down their stuff that they do like in the 19th century. I'm trying to get them to use electronic laboratory notebooks. This is, they don't understand what this is. This is very hard. If you talk to them about the blockchain, they will give you this. So I think the blockchain solves problems. A lot of problems that we don't even have. It's a promise, actually, because there is no robust application for the claim as you've seen, the guy knew what he was doing. This is maximum what he could do. And I think much simpler technical solutions and concepts like ELN's Repositories Pre-Registration Open Data and OSF is already there. The problem is it's not implemented because people don't want it. We have to tell them why this is good for us, what it can do for them, and so forth. It's not that we are lacking a technical solution. Now, I think in a way it's a bluff. It's a distraction. In fact, I'm almost sad about it that we talk about this because it's a real distraction. We could do useful stuff and not talk about it to claim that there is a technical problem. There is no technical problem. What we really need is we need to assess the science, not proxies, we need to talk about science and not about technical solutions. We need to read the output of our scientists in our communities. We need to read them, not to put numbers on them. We need to, if we want to assess it, we need to assess its impact, not in terms of its impact factor, but rather what is the impact in terms of in the field, advancing knowledge, creating new therapies, whatever. Scientists need to be more cooperative and funding needs to bet on more horses. So that was my rant. And so this is the Institute, besides being an experimental neurologist, I also tried to improve science, at least what I can do about it. And I'm running an institute that's called the Quest Center and that's at the Berlin Institute of Health. And if you're interested in what we are doing, you can check this out. And this is where my talk is. Thank you. Yeah, thank you very much, Professor Dinagel, for this nice talk. I'm pretty sure that many, many people think the same and we have a lot of stuff to discuss here, yeah? And thank you very much. I'm really glad that you gave the talk in this way. Cool, so are there any questions or we have like microphones here? Like discussion who wants to start the discussion, yeah? Okay, yeah. Hello, so I wanted to say just something a little bit in the defense of IPFS and those kinds of infrastructures. They're a little bit in their very early days, but there is another kind of a layer on top of it that in the future people could be using like Filecoin, which is supposed to incentivize that people will be actually using that system. And the power over this, in comparison with this centralized system where you have one cloud storage and you put your data there, is that a little bit like Torrent today, people could be mirroring and having those files everywhere, a kind of spread globally. And then if you kill it on one server, it's still there and people can still find it by this immutable address, not the link. So I think in that sense, it is a bit futuristic and it will make sense in the future to use it. Thanks. I understand, and basically I think it's a good idea and we should use it. I'm just saying this is not our problem. Just a question about the funding idea and having blockchain as a technology for getting additional funding. So I'm with you that we should use all the traditional ways or tax way, DFT and so on, but you might know as I know the percentage of successful proposals to DFT and others, even on the European way, are not very high. So blockchain and peer-to-peer funding on this way might be an additional way to fund real substantial research. Absolutely agreed. Again, why do you need a blockchain for it? You just said peer-to-peer. Peer-to-peer is the principle, but blockchain is a technology and I'm not saying that blockchain, you shouldn't use it, but it's not a miracle. I mean, the Kickstarter and stuff, I mean, they don't use a blockchain and you can get, I know people who got funding there. So yeah, why not use it? Nothing special. Hi. So you said in the end as a proposal of things to change, it's like the output should be red. But in the beginning you said 95% or 90% of the stuff is anyways wrong and you also mentioned it. No, no, not 95%. He said more than half. More than half, okay, fine, but still. Most, he said most is more than half. True, so the question is why would I read it? I mean, I got quite frustrated in especially the biomedical field that indeed you can find an answer to every question you ask because all directions have been answered. Sure. And the storytelling, I think, as I agree with you is really a problem. Because we don't read it, there's so much output. I mean, if we had to read it, there would be less output. So I'm for less output. If you incentivize output in terms of a proxymetric like either numbers of publications or an impact factor, you're no longer interested in what's in the paper. So if we start, and the problem of my argument is that I'm also like you, I'm kind of hallucinating. So this is not where we are. I'm telling about where I want to be. And I want to be in a situation where we read our papers and judge them on what's in there. And then we have less and less papers. And just one last remark also on papers not being replicable and not being true. There's one big problem in this reproducibility discussion and that is I think a lot of the stuff should not be reproducible because we are doing cutting edge stuff. We just need to be aware of that. This is the case and it needs to be replicated and then we see it's, we need to go somewhere else. But if everything would be replicable, it wouldn't be science. It would be just damn boring stuff. So yeah, the output has to go down. Yeah. Hello. I would like to disagree with you. And in my mind, there's a certain flaw in your argument in a sense that you kind of underestimate the appeal of blockchain and crypto economy in general in its role of a kind of ice breaker. What I mean is that not technology as it is now might be valuable and useful to scientists because as you said, right, most of the things could be done without blockchain. But it exactly has some potential to change the rules of the game for the funders as well. As in the last slide, you said that scientists need to be more cooperative and you need to get more basic findings. That's a very good will, good imperatives, but they may not work out as they do. But for example, when a certain governmental institution when you go it and say, give us more money, it says, no, you have enough money. And when it's the decision makers are presented with this idea of more transparent, more decentralized, or just another way of running science, they say, okay, we might give it a try. I mean, the scales will fall from their eyes. And they say under this, just for the appeal of this new technology, they might change the rules of the funding. No, I think you're confused. First of all, there is no appeal of the blockchain to any scientist or government funder. They think this is idiotic. It's a way where people speculate, make money. So this is an aversive technology. I don't agree that it should be aversive, but it is aversive. Second, you're confused because what you want, and I'm probably on your side in this, it has nothing to do with the blockchain. This is a different type of funding. Blockchain, I mean, you know what the blockchain is. I mean, this is a distributed ledger on a cryptographic basis. What does it have to do with funding? I mean, it's just a technique. Paper, I write my grants on paper and they review papers. It's another technology to do that same thing. What's so special about it? The only thing that could be different is this trust thing, like that you have a distributed trust in this thing and no central authority. Again, that may or may not be true even, but even if it's true, it's not our problem. Yeah, here. I have to give, yeah, I have to say that I'm here. We are doing social scientific studies and collaborations on scientists who do actually collaborate together and our findings are the same. So they are completely unaware of the tools that they should or could use. And if they are aware, if they are motivated to use that, they still don't use it because it's a completely different system. They are a different system for different grants and they don't want to use that at all. So we actually observe the same what you just said and it would be interesting to see if blockchain as a revolutionary idea could revolutionize, change the whole system behind it. But I'm also skeptical on that and it would be nice if you could surprise us that okay, yeah, blockchain could do that. But my question is that can you force in a kind of strategy that you take along scientists in the whole process and then change the way they think? Would education be useful to change the mindset and the whole hierarchical structure of the science processes that's going on now? Yeah, well, otherwise I wouldn't stand here. So I'd just be in the lab and write cell papers. But so I think it is about, certainly it must be about education because it's about changing minds. And what we need is kind of, it's a huge behavioral change experiment. It's a massive complex intervention that we need and we, when I say we, it's kind of many, many groups are trying this now and we're trying this at the BRH. And what it entails is on the one hand to talk to the young scientists and they totally understand this and I think they're on a good track. The problem, of course, is that they then run into the system and get kicked out. And then at the same time we need to change within the system. So we need to talk to our professors, to our commissions and so forth and we need to change the mechanisms there. That's something we are trying, but that is obviously the hard part. In the end, I'm still optimistic, but it will take a lot of time. And it will not, the key will not be any technology. And it's not against the blockchain or any other technology. There is no miracle fix to it. It's in our heads. Why would a technique change this? This is something in our heads. So we need to change our heads and this you do by education. So I think it's education. Ah, that's what most people say. Yeah, there are companies, technologies, or the likelihood of us aside just empowering ourselves and actually making meaningful change rather than having these little communities coming in to practice healing to us to come to make the actual change because presently on the side of this novelism, I don't have the time to look at this. I still need to get funded or I still need to be relevant in my field. So is it not likely that you have these little small communities coming in, cracking the ceiling to us, and then we can, like, first print what we really want to do? Yeah, agreed, but I think it needs to come from both sides. So because, I mean, you mentioned the funders. I think the funders play a critical role. And so one other, besides education, besides small communities that do something about it, the funders have a key role in this and there are some silver linings on the horizon. Take, for example, the Wellcome Trust. The Wellcome Trust does a lot of good stuff in its sense. Take Volkswagen Foundation in Germany. They have pretty crazy, but I think very smart mechanisms of funding. They have, for example, a lottery. There's a format where they just have too many applicants. That's, well, we have a lot of applicants. What do they do? They first start with a lottery. I mean, why not? So I think the funders, the change has to obviously come from many, many sides. As you said, I think the individual scientist is the weakest link in this. I mean, you have to, I mean, get food on your table somehow. Thanks for the talk. Yeah, thanks for the talk. I think it's really important. Like, we're all scientists here. We're mostly scientists, or we're thinking scientifically. So we need to think skeptically. We need to make sure that we are addressing the real problems and considering that everything that we're talking about might be wrong. But I think what you've been saying, people have been saying this for 25, 30, 35, 40 years. There's the idea that we need a culture change in science is not a new idea, and so we need to do something different. And I don't know where the blockchain is the answer, but it's clear that we need to take different approaches. Well, you're right. I mean, it's not new, and I just read a paper over the weekend by Robert Merton 1968 science about the Matthew effect and that there are too many outputs. And so half of it could have been in... The first paper was 1942. Yeah, I know, but well, I read the 68 one. So yeah, you're absolutely right. But nevertheless, if you look at it and it's way too slow and maybe I'm illusion, I don't know, I'm biased and I'm too much into it, but looking at the curve of what's happening, I think we are seeing change and maybe I don't want to be a party booper here. So maybe the blockchain is the answer. I'm just very skeptical about it. Maybe this is part of the thing. We'll see. Yes, thanks for the super interesting talk. One question. What about the vision of having not to rely too much on platforms? So my hypothesis is that we got used to put much of our trust as researchers, as contributors to research of any kind to certain platforms. And you mentioned in your slides the problem of proxies and assessing research results. And I wonder if this is not closely related to each other. So because we feel like, okay, we have to rely on lots of platforms anyway, we have this competition on brands of platforms and use them, misuse them as proxies to assess research results. And I wonder if this vision of moving research transactions to appear to peer network where those who actually have these kinds of transactions, for instance peer review, actually own in a very strict meaning these transactions and where you have a secondary information market that makes use of these information that are then available as a commons. I wonder if this is not well worth considering as a vision for research infrastructure. I agree as a vision. I would completely agree and lots of the other things that I presented were visions, so we all have visions. But it's an interesting comment that you make. Again, I think what many of you confuse is a technology and an intervention. You are talking about peer to peer and community and so, great, yeah. But you don't need a blockchain for it. I mean, the only point I'm making here is don't confuse the technology with a culture change. I mean, it may drive it. I'm very skeptical. I told you why I'm skeptical. But you said we're relying too much on platforms. I think you're on your way to relay too much on another platform and that is blockchain. Or it helps us to get rid of platforms. Thank you very much. Thank you very much. Ulrich was really great. And now we have some background.