 Okay, thank you. Yes, thank you, welcome back. I hope you are refreshed from the short break. And thanks again very much to Charmin and Zürich, who came up with all of this here. And I think it's great. The other day I had a short conversation with Alex about what seems to be some of the exciting thoughts of this area is that you have this complex mixture of things that you need in order to understand what happens. Yes, you have things like economics game theory, especially you have like applied cryptography, you have other areas of applied computer science, and so on, and so on, and so on. And you need, for instance, also domain experts, which is a whole different thing than again, like, for instance, from research you need insights of how scholarly publishing works, you need to know about how academic institutions and research and learning works, and so on. So it's fascinating, and every time you learn new stuff. So what I'm about to tell you here is something that to start with will get slightly meta towards the end because I gave a variation of that talk two weeks ago at the web conference in Lyon, and it turned out that at the same session we had Tim Berners-Lee there, the inventor of the World Wide Web, and we had a very interesting discussion on how blockchain might help or might not help with very basic problems of scholarly publishing and opening science. So towards the end I will tell you a little bit about the outcomes of this little discussion, which I find fascinating. But let's go right into the matter at hand. First, let me give you, because most of you, I don't know, you don't know me, let me give a little bit of a background where I'm coming from, yes? I'm not a researcher, but I'm an academic librarian. So I work for different research libraries since now, already 14 years or so. Around five years ago I founded a new group in the research and development department of TIB, which is a German National Library of Science and Technology. It's called the Open Science Lab. And what we are doing, just to give you two examples, I want to be very quick here, is for once we promote Vivo, which is an approach to current research information systems. So if you are working at a research institution that wants to set up a database where you can easily find out about all of the research outcomes of the researchers at your institute and you are currently pondering like thousands of other institutes all around the world about subscribing to, for instance, Elsevier Pure, which is a current research information system, think again, instead switch to the linked open data approach, which is Vivo, which is much better. This is not a topic for today, but just to mention the kind of stuff I'm usually into. And to give another example, have a look maybe at the URL given there, book.fosteropenscience.eu. We set up a book sprint by which I mean we invited 15 researchers from all over Europe who are more or less specialized on giving training and data science and open science and so on. And we wrote a handbook for people doing open science training. And have a look at this as well. And so you see my point is not maybe here and then doing actual research, yes, but I'm more from the research infrastructure side of things. And I was very delighted to hear this welcoming note today from a actual head of a library here, very good point to invite him over. And yeah, that's where I'm coming from. So the actual thing that I want to talk about today is a research assessment. This is only, as you know, one aspect of what this whole research infrastructure or actual doing research is about. But as you already know from Ingo's talk this morning, this is crucial, yes. Because you have, first of all, you have research assessment virtually everywhere, yes. You have this most atomic element of giving peer review, but then you have the higher level functions of committees who all of the time decide about things like hiring, like tenure, promotion, funding, prices, fellowships and so on and so on and so on. Mostly built more or less on this most atomic thing of the peer review. And as you know, this is only one aspect, one point of view on the research system, on the system of academia. But it's a crucial one because we have a huge systemic issue here, yes. And Ingo introduced us to this very briefly, yes. The problem could be stated in that way, but I say, okay, these committees that I mentioned just, they mostly rely on proxies to assess research quality. And one of the most well-known proxies is of course the journal impact factor. So instead of looking into the actual research, they rely on some indicator that tells us about the perceived, some perceived quality of a venue to publish that research. And this is really hugely broken and a huge issue, yes. And if you take away only one practical thing that you could do right now from my talk, let's check back with your research institution if you are from academia, if you already signed up to the San Francisco Declaration of Open Research Assessment, yes. Because this is a great way to involve your senior researchers in this discussion and bring this to their attention, that there is this issue. And I could go on with this, but I won't. This is just to open this, to think about this. And at the same time, this is not only a policy issue. It is a great policy issue, but it's not only on that level an issue. So why or how? My thesis here is that we failed to give researchers agency and transparency about research assessment. And as somebody who is involved in the area of crypto economics, you maybe can already tell by the set of my slide that I view it very much to the lens of crypto economics. Yes, I would say the challenge is to give researchers agency, by which I mean direct control about their identity, their assets, which is the actual research outcome, and the interactions between them, which is when we talk about research, not a currency or a monetary credit, but research assessment. This is the issue with research assessment through the lens of crypto economics. So let's have a look where we are with this, how we failed so far to give researchers proper agency. One thing that comes to mind immediately is ResearchGate. Who, by the of you, has a ResearchGate profile? That's nearly 50%, maybe more. That's not very surprising, given the huge success of ResearchGate. There are literally millions of researchers there. And on the surface, it's very nice. Just the other day I had a talk about this with Sönke, where we recognized that this is another iteration of something we have gone through maybe 10 years ago or so when we both, but of course many other people as well, promoted open science and science 2.0 and stuff like that. And these were one of the champions that came out of this revolution, yes? On the surface, it's very nice what you have there as a researcher, because you are in the driver's seat, yes? You can decide, okay, I want to state some attribution that I gave to a work of research. I want to give credit to other people's work. I will show off my network. I will see what other people tell about my work or our common area in my timeline. This is really valuable, and it puts the researchers in the driver's seat, as I would say. But now, what's the problem with this approach? There is this one company in Berlin called the ResearchGate GmbH, and they own all of the data. It's literally in their terms of services that every time you give up this valuable piece of information, it's like a present to them, and they are allowed to do whatever they want out of it. So have a look, for instance, at the most recent agreement between Springer Nature and ResearchGate, which essentially says whenever you put up a paper there and you're doing a major thing for some reason, that copyright does not allow you to upload it there because it belongs to that market somehow, they can pull the trigger. This already happens like millions of times, so people, a paper, just vanish from ResearchGate. This is usual there. Or see, for instance, just as another example, the ResearchGate SCORE, yes, which resembles this one-dimensional way of doing research assessment as we are used to it from the age of journal impact factors and so on. It's very one-dimensional because there's only one company in control in place, and they set this up. It's utter bullshit. Ask anybody who's into bibliometrics. They will tell you why. I won't go further to this, and it doesn't help in any way. And this is just to mention two of the examples of the problems with this model. And then have a look at another model, and that is Orchid, yes. Who of you in this room has an Orchid identifier? Ah, that's great. I love you. You are a great audience. So you are a very literate researcher who knows what is going on and how to behave. But the same with Orchid is, it's the best what we have in the area to disambiguate between researchers and their names and to attribute them properly in one place. So it's one centralized open database of all the attributions that we need. Again, the idea is fine, and in many ways it's a counter model to what I showed you just before, research data, right? Because it's all open, and it's not controlled by one entity, but it's very properly governed by consortium of large publishers, libraries, and so on. But there's still another problem here with this approach, and that is that Orchid is by its nature largely an aggregator of research metadata, so to say research information, which is produced and controlled by big publishers. So they have a huge systemic issue with metadata quality, because nobody is incentivized or even allowed to directly control what is there, because it's all through the proxy of data which is produced and controlled by big publishers. So again, another systemic problem here. Have a look at my article at the London School of Economics Impact, if you want to learn more about this kind of issues. Now, if you think about it, I need to be more quick here, but this is a problem of statement and maybe already the most important slide of my talk. Look at it. We have an issue with data ownership here, yes? Or to rephrase it, a problem with the agency of those semblance and perceptions of data. And what I want to point out very quickly now is that in another related area, which is educational certificates and higher education, people already two or three years ago came up with a solution to this kind of problem. And that is called block certs and it's a blockchain backed educational certificate where you as a recipient of some assessment of your work ask somebody on the blockchain, which is your research institution, your teacher or whatever, to certify you or to assess, to certify something you do. And then it's completely yours, yes? It is immutable, it stays there, and you decide what to do with it, which is a great thing, which really is a showcase example of what blockchain is about, what it has in store. And I will not go further into this because of lack of time, but a very important point, the technical concept which is discussed in this area is self-sovereign identity, but as Charmin pointed out, and this is hugely important, this is essentially about individuals and their personal data. This is what is at stake here. And what we have is a decentralized identifier as an option for self-sovereign identity. And although all of this, I would say, what we have is blockchain is still experimental and somewhat hyped. This is a very hot candidate for solving some of the problems with data ownership. Because the thing is that claims on such systems are cheap, but they are not for free. And this is a huge solution for economic sustainability of a system that is really not owned by one particular party that is involved in this. As you know, this is so special about this. And the other thing is it is virtually publicly owned and only you are the sender, you are the party who is responsible for what is said, what is done in your name on that system. And so this is more or less my last slide. The idea would be to let every... And this is hardly new or hardly original. We already have, as you know, a number of startups who are actually working on this. I just want to have a common denominator or bring up the concept of what this is about. And this is that you let every researcher make statements directly without any detours, without having to rely on any particular architecture or server or so. And by the way, you fix the incentive, the broken incentive structure of assessment and metadata quality by this intermediating this process. And by the way, you also level the playing field for business model innovation. Which was called back in the 90s when Tim Berners-Lee came up with the World Wide Web, which is an innovationless innovation. Because you have a system then where nobody would have privileged access to this valuable scholarly metadata. But it's a common good. And just very briefly, but I will really end in a minute or less. So one of the interesting things that we have in this discussion with people like Tim Berners-Lee about blockchain in this area is that they assume that our protocols, and as you know, blockchain is all about protocols, not delivering these islands of vertical solutions, but the protocols that are open to everybody and which everybody can use. That this level of protocols is thought to be separated from the level of economies and incentives. Very interesting, think about it. It's maybe one of the things that is holding people back on the level of understanding things, an issue. And another thing is this issue of Ethereum or things like that, coming out of fashion. Also very interesting. But I don't need more time for that. And then if you want to have a closer look at this article where some of the ideas from these slides are further expanded and yes, thank you. We can allow at least one question or so. Two, yeah. We started off by talking about linked data. What have you come across in terms of the intersection between linked data and blockchain? Just very briefly. The overlap is huge and we will depend. We are not even started yet. We're doing linked open data right. We will need it because this is another layer. And alone at this web conference in Lyon there were two entire tracks only about the overlap between linked data and distributed ledger. That tells you something. For instance, just to give one quick example have a look at John Domingue and what he is doing with educational certificates at the Open University in the UK. They have all of the semantic layer of the Mozilla open badges where you refer to a learner's e-portfolio and so on with all the nice semantic stuff but then you have this new level of crypto economics. So it's not mutually exclusive but it really belongs to each other. Sorry. It is more about in the future when researchers do run the wrong data. It's a very certain school. How will you fund it? The nice thing about blockchain if you think about it is that you don't pay by having your institution paying a hefty subscription fee to some huge thing like a publisher. Neither do you pay with your personal data by using services for free from some big company but instead you pay a very modest fee to place the actual information about you and what you have done on a blockchain. This is the whole difference. This fee may just consist of a tokenized piece of resource that you have and that you abundantly have. So for instance you could say I have one gigabyte of bandwidth that I hardly use and 10 gigabytes of storage that I actually not use on my notebook I give that to the network and in exchange I get that token that allows me to do proper research data management attribution to my research and so on. Good. Excellent. Thank you very much. I like it. It's an untalk at an un-conference. You introduce yourself. Thank you. Hello. My name is Dan Quachat. I'm coming from basically two teams. One is Codex. We are a fresh startup in the field of crypto whatever and I'm also a professor at the University of Maribor which is not far from here to 150 kilometers neighboring country and you get here very fast if there are no traffic jams on the highway to Vienna like today. And so I would very much like to thank first for the opportunity to organize this and we exchanged emails and telegram stuff and it's really nice to see that there is an open community building up around these topics which for science are quite important. And also I would like to thank all the previous speakers to give a nice intro into the field so I can skip a few slides and I would like to make just two points actually. First I would like to really connect blockchain and science and see that blockchain technologies are really the perfect match to solve some of the problems we already identified this morning. And secondly is to give you a brief idea of what we do. So why are blockchain and science a perfect match? If you look at the idea of the blockchain that you have three types of decentralization going along in this area. First is architectural so there are no three points. The next is political, there are no agents centralized who make decisions. And the third is logic. So there is one common degree of state and if you transfer this to science it's basically the same thing. And science is actually the only social process that we know of so far and it actually leads to a broad share of consensus. And this consensus is formed around very good statements that are true in scientific sense of course. Now, not everything is rosy as we discovered before especially in publications incentives ranking or whatever. So something is seriously wrong in science and not just that people are trying to patent mathematics lately but also other stuff. The process of publishing, reviewing citation carbaz unit. So also we like to forget in science. So you have papers that have been retracted but they are more cited after they have been retracted than before. So basically this is a good example that this stuff with publishing doesn't really work. So if a paper is flawed and then retracted and nobody knows that and everybody still cited this paper so you're citing something that is inherently wrong and it still goes on. You can look up the paper as both citations and still cited in 2018. It's like everything is okay. So you need something that would serve as a chain of let's say events, history that it's not for them. Okay. So the basic unit that we use for communication in scholarship or academia is basically the paper. It usually is pretty obnoxious to read this stuff because it's written in some strange language bordering to English but not quite full of symbols, graphic cons, nobody understands because it's a jargon. So there are two nice papers in Atlantic and it's really absolutely both targeted and also all wrong what's the future of scientific paper there more into that we need to have open code, open data with the paper so we can actually check what the stock is in there. But of course this is not the only answer. What we think which is our core idea of our solution, tomorrow we will have a demonstration, we can talk to our CEO sitting there and some other guys who will be from here and we can show you a few demos and stuff how it's working. So the basic is that publication itself has to change its nature. So two of the authors are actually sitting here on this idea published that the publication should be dynamic and collaborative. So if you condense this message into a few keywords it's basically future paper should be dynamic so it should change all the time it should allow remixing so you can take something from one paper so the core idea mix it with the other, put it in another paper should be collaborative. So three pillars but do we have the technology to do that? The idea is that you can look at the paper as a contract first. The smart contract because we are on blockchain conference and basically our technology that we are working on is of course the Ethereum but before even it's a smart contract if you think clearly it's actually a contract and it's a contract between the authors who has to agree that what they will publish is actually true if they are of course true scholars otherwise they are frauds and they are nice also like that so you cannot fight that that's impossible so there should be a contract between them it's not a legal contract a law contract it's a social contract actually and the research paper is actually a contract between this group of authors and the community that they what they express there is something new they bring something new to the table in science and they stand and they take a stake for that. So how to do that in a few coding slides if you don't mind so that we get some nice pictures also you can think of a paper where it's coded into something as a finance state machine so a paper has a few states and a few transitions between the states if these are carefully thought through you can actually optimize this stuff you can actually come up with a paper that you can transfer this finance state machine into a smart contract basically instantly and you can draw a transition graph so in this small example for instance as an author starts with an active paper so it publishes introduce it for instance into our platform and then he waits for some community response if there is no community response well obviously it's not a good paper if there is some community response either way or positive this state for instance transfer into a recognized state so the paper is now recognized by the community and it could go through technical report so there will be guys who will be interested and they will write a positive or critical or whatever report and they will take also the state in writing the report naming by being uniquely like an identity so no more anonymous reviews in which you can hide behind your unknown identity you can write whatever you like you can steal ideas and then in a few weeks you will know this stuff so from recognized paper you can have positive technical reviews negative technical reviews you can hand out in indoor state or you can hand out in a locked state but that's not the end actually even if your paper is locked for some time if you still get positive reactions you can unlock it and start the process again so the paper actually never dies of course this is a very simple graph what is missing here is the economic system that drives this stuff so the incentives why would I do that so obviously I need to take some steak if I publish a paper so I need to put some of my tokens someplace else and if I will be successful in publishing I will get some more tokens back in the end if I try to publish some rubbish I will lose these tokens so I won't get to try again there should be a clear economic incentive not to overpopulate the unpopulated scientific publishing field if you believe in your ideas you will fight for it you will take a steak you will take a chance that you will lose something but in the end overall if you are really true if you are really a scientist you will be winning again okay some idea that basically this type of stuff is already running so you can code this state and transitions into simple Solidity functions and since this is a smart contract it also doesn't die and if you now take a few steps into the future you have a blockchain system platform whatever and you have myriad of contracts going on in the blockchain but basically what you have you have an autonomous entity that lives on the blockchain and other papers can interact so other contracts can interact how, for instance, by citing each time you cite something you are interacting with a smart contract you either make a transaction to that contract you make a call you make a message so in a landscape of different entities that live on a blockchain a paper has a smart contract to be something between distributed or decentralized application and decentralized autonomous organization and if you take a few contracts together basically you will have a journal if you have many contracts together you have many journals in one time, in one particular moment of time there is a journal when there is an interesting topic otherwise there is no journal and the journal will be just a selection or just ad hoc collection of papers in the same field that have very narrow focus so all the people who will read these journals will be interested in that papers today there are basically no journals I mean as a practical scientist who actually reads journals I don't maybe I'm a minority but I never read journals, I read papers and I don't care where these papers came from if they are not on the internet if they are not in archives or pre-print servers I will not read them anyway and if many of these scholars will do that well there will be no journals anyway but that doesn't solve the problem of publishing because there is always a problem of trust and trust and incentive is what is basic to blockchain technologies and what is changing the way people think about human social systems so basically this is our starting site we can visit it today, tomorrow, show it to your friends spread the word ask us anything we have to cooperate with other projects in this field there are many projects in this field and we would like to contribute to this community actively basically what you can do you can claim you can claim a paper that is already in our database you have to install a metamask wallet for now we are running this on a test blockchain public yet but it will be soon so you can see how the transactions are when you claim something you can review it there is also the interaction between the entities that claim from the blockchain and your article is on the blockchain it has a hash there is my Ethereum address there if somebody is interested to donating something something and what we like to do with this project is actually we like to help those researchers who are unable to publish today to express their ideas and even to review why because they don't have access to this system they are probably living in a not so well off economy they don't probably they have very difficult access to internet but I'm sure that these solutions are the step in this direction because the talent in the world today is much greater than obviously not so we have half of the world that doesn't have internet access at all so now imagine a few years when these numbers will go up we will have really all talent at our disposal to advance science thank you I'd really like to talk so forgive me if I try to answer that so we looked at the history of publishing universities the incentives and how they corrupt good aspirations we looked just recently at the research game and it was great business model that's the problem I frequently try to collaborate with you but you have a certain legal structure and your own incentive structure which can make that difficult to collaborate with you so obviously as far as I can see you will make money from the actual smart contract itself and I could just fork that smart contract and do an equivalent version of that but I'd like to add those smart contracts and open access journals you're working on have you looked at your own legal structure and the way you're financing and examined how that might steer you in a good direction or a bad direction and for instance in particular why would you structure yourselves as I don't know how you structure yourselves as a private company rather than a co-op or something like that I mean the structure the company has many advantages because you are your own boss actually you can have access to different types of funding and other stuff so you are more responsible even if you are in a small group and run as a startup or as a community but anyway our idea is to have a public blockchain in the end so everybody could publish everybody could review everybody can read everything everybody can set up their own identity in the platform there's no problem with that the only thing is that you can actually earn as an individual you can earn by reviewing and by publishing which is now you are and I don't do that even if I get an invitation I decline actually because I would not like to support this kind of free work of scientists even if it's as a community service but I'm not serving the community I'm serving a huge publisher who claims the copyright afterwards because you are for instance in a company a piece of transaction piece of every transaction that happens in the blockchain but you as an author you earn much more for instance you have a couple of tokens issued and you would like to publish a paper you put a paper on the blockchain if it's successfully published you get a few tokens more back if you write a review if this review is positively judged by the community it can be a negative review but it should be viewed from the community as something contributing positively to the community so you actually found a flaw and a paper or something like that again you will be getting some reward in terms of crypto economic incentives it's actually a network if you have enough people the system works if you have two people it doesn't work the only thing yeah I think what they want to think of though is how is that going to help you with the motion you are right you are a professor I think you are not how does it help you with the motion if you are more connected to my scientist but no no no if you are still starting out why would you do that I'm just asking if okay the incentives how many kids have you paid a little bit really in academia because the goal is to get commercialism this is the problem with everything you do in this field because if this incentive doesn't change so but established researchers professors who have gone through the path they are probably not interested in this stuff I am because I don't think the system is fair because I can see young people leaving academia for the wrong reasons but as in any other social incentives or systems things doesn't happen overnight so if you would look at the papers from I think 1995 or something there was a famous paper that what is this word right for instance in New Suite this one word you know what can I do there I cannot do anything there so basically you are exactly the same spot you have even huge economists and computer scientists who are against crypto crypto field or whatever because it's definitely reshaping the field so if you give the incentive to the individual many things so but it's not easy but if you don't have any other option if you are outside of academia and you want to publish you want to do something you want to review you cannot review nothing nobody will invite you to review something but you can review here by yourself you can write a review and publish it if you have not done anything before and of course if you are doing positive things you are rating your recognition in the community to your rise maybe some of the bureaucracy maybe some of the influential but very quickly one thing you might want to do is to get people technical experts to review journal articles because much of it because the reviewers don't know what it's actually about they try to not know if they really overcame actually you can claim four papers and review them yeah that's the idea thank you very much thank you Marius from ScienceFood to talk about well it's the title set right you can just like introduce yourself if you want so let me first introduce our founders Vlad Grimtor he has a broad knowledge in IT finance and consulting he has worked for major companies like Bosch and Microsoft and Alex Kirica he's also my brother he's a physicist he will start very soon his PhD in molecular dynamic simulation at the University of Vienna and I'm Marius I hold a bachelor and master's degree from the University of Vienna and my background is in experimental particle and nuclear physics ok so I'll be talking about we will propose a new way of scientific conduct and how can we incentivize the community with blockchain so I will go firstly through the problems very briefly so number one I would put it's limited access to research output research papers are not free and do we need open access is the question well I would say yes because in order to create research we need access to information what is creativity it's taking known elements and put them together in original ways Satoshi didn't invent bitcoin the elements were there he just put them together I have a friend he wanted to be a grandmaster in chess by using the power of mind by doing medication it's not a joke but unfortunately it doesn't work this way if you want to be a grandmaster at chess you need to study the games of Kasparov Alehi right in a similar way if you want to be a great researcher you need to study academic research papers and in order to do that you need to have access to them and yes we need open access there are no rewards for the authors or reviewers this leads to a slow publication process the reviewer part there have been discussed the reproducibility crisis and all the successful results I've published I won't insist too much on that but the last part is as a scientist you get very often stuck in your research and it's important that you have a forum where you can ask questions and get a response to it ok so what we are proposing at science route is a scientific ecosystem that incorporates all the functionalities required during the scientific discovery process and this is a collaboration platform funding platform and a publishing framework ok so I'll go through them I'll start first with the collaboration platform we'll have unique identities and profiles we will have an incentivized questions and answer forum which means if I have a question and you answer me I can pay you in form of science tokens I can tip you this will improve the collaboration we will have a search engine for papers also another aspect that will integrate is a decentralized marketplace let me give you an example what do I mean by that if I have made the nanoparticles and you have the microscope right but I don't have the microscope I can give you the nanoparticles and you can look at them and by doing this without the third party you can improve collaboration ok we will have the concept of scientific repository which was discussed already so what happens now we only see the publication we don't see the whole research process we want to integrate the whole research process into the everyday life of a scientist how would that happen I take the data today in the evening I run it through the hasher I take the hash I put it on blockchain it will be placed in a decentralized storage system but now we don't have this decentralized storage system so we will use normal storage system and we will use blockchain only to timestamp to have this proof of existence right it will be like a github for science where scientists can fork ideas and don't worry about knowledge there good we will have a funding platform which will integrate an index of international funding system and opportunities we will have this idea of crowdfunding which means if I have a great idea let's say I want to raise the 1000 euros for my microscope and the community wants to support that they can do that in form of signed tokens directly without the need of third party implement smart contracts to it let's say if I want to raise 10,000 euros in one month then if I reach the 10,000 euros then the 10,000 euros the money goes directly for the microscope if I don't reach in 30 days this 10,000 euros the money goes back to the investors blockchain grants will be easier to configure and distribute okay we will have a rewarding mechanism for institutions bitcoin and ethereum they work and they are decentralized because there is a profitable to do mining that's the reason why they are decentralized we also want to be decentralized which means that we will especially encourage institutions to secure our network through staking codes we will be proof of stake and we will do that it should be profitable for them this is the only way we can achieve decentralization okay and they will get reward for staking coins so first they have to buy them stake them, secure our network and they will get transaction fees this will be also part of our business model we will stake some coins and we will validate transactions and we will take a small fee okay we will have a publishing platform which will be a framework for journals yes so this is the current publication system so the author sends the article to the reviewer the reviewer to the editor the editor to the reviewers and this whole process can last more than one year this is a problem with data security you don't know what happens with your data during this year what we want to do is to create a timestamp monostip submission so at the moment that you submit your article to these journals it will be timestamp and you can prove that your data is yours and we also would like to make a blockchainified record of every peer review this is a way to incentivize reviewers so basically what happens here already established journals can come with their own business and take this advantage of timestamp monostip submission and this proof of prestige why is this important we would really like to make it a requirement for journals to provide their authors with this proof of existence we would like to do that okay and the last thing we would like after we establish the community after we have the blockchain ready we have the scientists everything is ready at that point we will have our own journal so not before, we need to build the community first no one will publish if we start directly with the journal and we will be decentralized and open access we will publish some of the negative results we will have a rewarding system for authors the website that authors will be rewarded and a rewarding system for the reviewers like paid in money they will be paid depending on the number of pages that the article has so regardless of if the article is published or not reviewers should be paid because he is the expert evaluator right okay why is this proof of existence important let's assume let me tell you why is it important so if I make a breakthrough discovery let's say I find immortality which is not so hard to find if you think after some age the cells don't divide anymore as they should if you find the reason why you might leave a thousand years this is a breakthrough discovery and well I wouldn't be I would be very anxious to publish it to send it to publication right now to be honest I don't know what happened with the data someone might steal it and with this blockchain you create this proof of existence and this is our main objective okay so the milestones to sum up what I've been talking about you'll have this social network it will be like a research gate that can showcase their work we'll have this funding platform with academic jobs and idea crowdfunding and we will have a decentralized publishing framework with unique rewarding system for reviewers ScienceRoot is not the work of a single person it's collective work of a team and we would like to give a special acknowledgement to conclude I would like to tell you a follow-up thing washing hands was a subject of debate some doctors linked washing hands with spreading of disease but it was a subject of debate with the invention of a microscope where they could see the microbes at that point we had general acceptance among the population with these hands we want to do the same thing with blockchain in the scientific community but for this we have to find this microscope for Bitcoin I think would be scalability for us we I think we are already scalable and we just have to prove that it works so thank you for your attention from what? we will have a charge at an acceptable price we will have to do some research before this will be the last thing that we will do nowadays there is this boycott Germany against Elsevier I think you heard about it and I think the negotiations are not led accordingly it's a very primitive way of negotiation because think about it they claim the Germans claim that open access should be at an affordable price as if it doesn't do that and they cancel subscription fees I believe it should be for free but how do you incentivize the editorial board and dream yourself if it's free you cannot put 4,000 feet on an open access article maybe Germany could pay 4,000 but how could Poland or Hungary or Romania how could they publish block science even if you have some research and decide what is the price you have your own token in which you pay the fees and it's highly volatile until a huge wedding and this wedding is more than 100 years old if you make it unvolatile and once it's volatile you have to have a mechanism to govern this and this government should be done by the community and this community should also be incentivized in both two sides those who pay for the reviews and those who receive to participate in this government and to conclude on a fair price so it's not because it's the price what you said before the price the price is going to be set and you will pay to us to deal with that but then when you still pay for health but from both of you yes but in general if you are using from outside you don't know the price you know but it can be mean to the price because token can be very much under to many states because it's free to people so that's why so I would suggest if you want to introduce me to you there is a concept which is called smart pocket which is paid by which is the percentage so if you with some money you can use but not in that way use the tea it's a problem it's just yes so you have the whole thing based on tokens I guess you could use decentralized exchanges and all that kind of stuff but if you're staking tokens for the for validating transaction yes so why are you not able to then help govern the blockchain I guess what I'm asking is with the network itself not determining what the price is what the procedure is is that not flexible based on the governance model well if you really want to be decentralized I think you cannot at some point you cannot decide no more like to change the rules I guess I'm hesitant to say or to agree with that because would you not want the community, like the global community to decide what the work is worth and then be able to change it how will you change it if you have 20,000 miners I mean validators how will you change the rules yes it's just governance it's actually rather possible it's kept on software every time even can't be made without software but there also is an issue which is it's cool idea and definitely should be done but the issue is that if you have a proof of state where it's really state so it means that those who owns most money will decide where the platform goes and this is huge part of securing our network so we could communicate with them do you know Chandler Goh? he's the main miner what if he comes and buys in the secondary market and then he will decide where the platform goes and we'll introduce proof of work there to move all the mining power to you yeah I know we cannot go with proof before because I don't see institutes no I just said that if somebody will earn enough of state he will decide in the future not you but we have a lot of are you keeping up 50% we'll sign 60% of the tokens yes but it's possible if someone buys all of them but I don't think it's either going to be you guys I would say is what it sounds like well ideally you want to decentralize so then you worry about government's problems but then you talk about hard forking soft working to get the right answer we have to have a good start and try to keep with that I think it's a good idea a good presentation you're talking about the whole scientific ecosystem so it's complicated of course why don't you just impose an ownership cap so nobody can take the majority that's a good idea excellent and I really love this conversation and I want to remind you for tomorrow when we showcase all the platforms and then we get really into these discussions and continue it like that is that okay with you good, cool so we continue with two more talks and hi guys how are you doing still some energy left yeah my name is Matias Bröder I'm a musician by training musicologist later on I was in the scholarly community for a few years and now I managed the estate of one of the top selling musical artists of all times so my main focus is business today I'm very thankful to Sherman, Alfred and Zernke for having invited me I'll be speaking about something that I find super interesting and I'm very excited to be sharing you know my ideas and then hopefully getting a lot of feedback from you guys this is an awesome community so about last year in December I came about this concept of token curated registries how many people in the room know of token curated registries not a third so I'll go over the concept real quick you can see a couple of links here if you google them you'll find it directly this is where this was first published I believe the person who started this discussion was Mike Golden and then there's also Simon de la Ruvier they both work in New York City for consensus or with consensus systems so token curated registries basically like lists and whatever goes on the list goes through a review process and the review process is very simple you apply to the list and then people review your work and if they like it you get on the list if they don't like it they start a voting process where both sides put stakes in and whoever wins the votes goes on or stays off the list so basically you can think of it like things for instance like the top universities with the best ratio of student fees versus later debt or something like this and universities they want to come on the list parents want to see that list and the people who started the list they want the list to be high quality so there are these three groups the first group are the consumers those are the people who really want to see the list, have it well curated have great content on it and they don't want to waste time doing all of the research themselves then there are the people who want to be who are token holders who have a token, who have a stake in that list these are the guys who really have an interest a financial incentive to have a high quality list and then the applicants or the candidates they are the ones who want to be on that list so they have an incentive to be in front of that audience of customers yeah basically this is what I just said so I'll go through it very quick so in a token curated registry you have application review and challenge and then voting, those are the central mechanisms so when you apply for TCR you basically have to stake token of that TCR that's the whole point there so you have to get that token you probably buy it someplace and you have to stake that token that sort of goes with your application it's like an application then once you have applied, put your application in then all of the token holders are notified that there's a new application and then there's a process starting and by the way all of this you can go on github take a look at the repository it all has been coded already you can start using it and so basically when you apply there's a time that is running and within that time period others can review your application so if someone thinks like okay this application should not be on the list then they can challenge it and they challenge it by also staking token okay so if no one challenges your application you go on the list and the token that you put at stake is yours so if someone challenges then there's a vote and the vote is secret and it's also secreted by token holders so if you own a lot of tokens your vote is basically has more value or more importance than if you have just started so whoever loses this vote will lose their tokens this is the basic concept it's easier now when I learned about this I thought it really interesting I was immediately thinking about a problem that I never really understood why is peer review such a pain why does it take so long why does it incentivize the wrong people what about all these stupid group dynamics in the scientific community at least the one in musicology where I was going so I thought hey this could be a cool way to use that so here are three user types again right this time around the token holders are like the primary caretakers of the germ they own the germ in a way they are a little bit like publishers or the owners of the publishing house you can think of them also as editors or publishers so it's a group of people who are really dedicated to this germ and they also include past authors so this is how the community of token holders gets built the candidates are those researchers wishing to publish in the germ they would like to receive credit and recognition they want to see their research paper on the list and consumers are also researchers those who are interested in the topic that is for instance the Journal of Computational Musicology it doesn't exist yet we can start it this way is this okay so far now I basically had very little time to work on this and think about it and very few people to interact with so I'm really looking forward to your feedback and discussion later on here are just a couple of things that I found interesting and that need further discussion so I think that if we use TCRs we can speed up the review process a lot because basically the default is you know looks good looks fine we did the basic fact checking the data is consistent we can reproduce the example the research so yeah that's a goal right then this you get much faster than asking a gazillion peer reviewers what are your opinions on this research paper please write a three page response so I think it will be faster I think that the comments that reviewers make during the review stage I think those are super valuable and in most peer review processes they somehow get lost they could be published here as part of the voting process I think this is a tricky one because if you are already an author of that journal you own token of that journal right now you put in an application for a new paper it would be you know not correct if you had the right to vote in that application so that's something to think about so essentially the groups that are incentivized somehow in this peer review idea they're going to have a lot of overlap or some overlap and I don't know what that means for the incentives and the model does it still work or not so token holders I think we have to have something like know your customer right it's kind of important that the reviewers unknown and that the token holders unknown also so if I apply to a list I would like to understand who are the people that I'm trying to associate my research with here who are the people deciding on where this should be included one thing that I find super exciting is that if you follow this model that essentially the journal is owned by those who produce the research or those who are token holders from the beginning and there are various ways to model that obviously but I think it's very powerful because then we get into the discussion of the price if I subscribe to this journal what is the price for that and that's essentially the community that decides on that and of course you have to factor in things like okay if we wanted to have a print journal I don't know if someone wants that but we have to have someone who makes a nice looking pdf and things like that so these people should be paid you can pay them and you can set the fees that you want to have for your subscription I think this is for me this is the most important part you own the journal that you are interested in the research that you are into that the infrastructure cost can be reduced from for instance the four-feeded tokens from a challenge right you have all of this capital that is being exchanged tokens you can set the amount of fees that should be to the network for certain things one thing that I like is the idea of you know how I see one problem when you say okay the default is that we let everyone on the list if there is no challenge right well you can think of a world in which no one actually reads the articles so one way to get around that is to say well if you own token and you are not participating in the review process then you know with every time there is an application your token value devaluates essentially so that over time if you are inactive you use your stake in that journal which is also you know a fair representation of what's actually going on you lost interest in reading all of that scholarship right so I think there are tools for us to make this really work lots of experiments needed also differences in disciplines right I talked to the you know my former grad students and friends who are now professors in various disciplines like in musicology everyone hated the idea no but we need to have our positions as editors and you know and then economics people also they were like hmm don't know if we need that so best results I got from the sciences you know so I think there are differences in the disciplines that you can essentially fine tune the token curated registry however you like so again take a look at the github repository and the documentation there I think they are doing a really great job I'm not a coder so I'm not participating in the coding okay one thing I have two minutes one minute you can airdrop for instance in the beginning to create the group of token holders okay so this is it questions, comments, criticism everything welcome yes David right one question okay thank you for everything I haven't actually looked in detail in this issue my background is in Intelligent Biology so when I look at that you've missed out to me one huge thing you know if I was to play this game let's actually ask the US election I would definitely have backed Donald Trump not because I agree with him not because I think what he's saying is true but because I would thought other people would come back in as well so what you're talking about here just as an isolation here it's a shelling game in which you're trying to predict what other people in the community will back in the doors on the list without any verifiable reference in truth so that in evolutionary terms is a very powerful isolation I might remark that like Simon's work is very very new and you have a bunch of mathematicians thinking that you can solve many kind of problem by putting everything into a code but there is like I think it's very young TCRs the idea of TCRs it's token curated registries it is a tool with which we can have future reputation systems but I fully agree that it probably needs a lot of evolving because it's still too economic games oriented and the incentive systems have to be adapted to this is what needs to happen now one remark first of all yes that's true it's a great remark the idea is that if you have a vote after a challenge that you also provide comments with it and those will be published together with your real identity on the journal and so if you vote if you vote for something that obviously doesn't have good quality sooner or later it will transpire that doesn't solve the problem of people doing votes of tribes to associate there are clear ways in science to solve this problem that's a good starting point that you need verifiable evidence feeding back incentives over the long term and that's what evolution does it's well studied, it's good velocity of knowledge in it and that by itself there is no question maybe you can point me to some resources thank you maybe one last comment I just wanted to your comment because I think talking about risk-trade is definitely a good idea but in terms of blockchain community how it should be designed well but you know I don't think it's just because talking about risk-trade is too young it's just because the whole industry doesn't really know how to design crypto and talking about risk-trade it's just specific too and it doesn't mean that it will solve all the problems it's really what I've also been experiencing 8 years ago where the system started to be the right and a lot of programmers were trying to put in which we are exactly designed for financial world into other type of domain and it will not work or it might work but it's much less efficient to see if you design it specifically and talking about risk-trade also might not work for all kinds of cases and you might try to tune it a lot of times to achieve something but it's better to start from this page because it's good design it's well-gained to our economy but I don't think that it might even don't feel completely one specific domain problem because when you start from the technology not from the business side the main side the social economic process you should start always from the solution not from design and technology from the workroom I mean my response to that would be that it is beneficial in a general systematic way to develop building blocks that can be used in more than one specific context and so I think token curated registries are more like if you say crypto-economic building blocks and they have to be actually tuned to different sorts of scenarios but I don't think that they will be closed than token curated registries so I think it's a good idea this was the starting point this is really like this whole discussion I think started in September of last year so it's still quite jumpy it's work on attention to our consumer markets is older the TCR itself since September I think it's a good starting point for solving the problem of reputation and curation and this thing but it's really just a starting point because it's super raw and as such it can be an element for a lot of the solutions we're talking about these two days but it's all very wrong so yeah thank you guys last talk just before raise of hands we made a reservation at Nikke which is a restaurant just across we made a reservation inside where 20 people can sit who would go to a joint restaurant okay so I think we will give you some instructions for later because on the menu the food will come earlier and then we can they have good food so to the last talk thank you very much hello everyone my name is Valentin I work with Iris AI and I'm also a scientist actually a physicist you're already a physicist by training so from first hand experience I can tell you anyway we've already had the angle nicely explained to us how the strange works system of science works anyway here are some of the biggies that we think will take science today or somehow at the heart of it all or most of it it's gross misalignment of incentives that pressure scientists to publish as much as they can which needs to reproduce the issues because they tend to oversell the results to get into highly cited journals that really work we have built in biases based on impact factors also of course a huge information overload because people publish a lot and people have huge troubles actually just keeping an overview of the field that is relevant to them and of course most of it is hidden behind paywalls so we have access issues that people that originally paid for the research I mean tax payers don't even have access to what they pay for we these problems and some others actually cost me to many other young scientists too and in my case to join IRSAI whose long term goal is actually to go ahead and fix these issues and what we do at IRSAI is build smart AI based science assistant tools for scientists to help them in the directory parts of everyday work and the tools that we have so far already addressed the first issues here information overload biases by helping scientists with the research discovery and review phase so not really have much time to this it will be there more about our tools tomorrow but imagine you are a scientist you have a brilliant new idea of new research project first thing you want to do is you know you want to see has this been done before has something similar been done before is it even possible you need to know to set out the mind so what you do with our tools is you actually start sitting down and writing up a problem statement in your own words like you would explain it to your co-workers for example and then in the first phase our exploration tool would then bypass any weird keywords that you have to assemble in a jargon that you might not even know for Google scholar and the likes it will actually read your problem statement and extract the ideas and topics and concepts that are in your problem statement and it will find relevant research for you based on what's contained in your problem statement it will present it to you in an easily navigable 2D map that is pre-categorized in automatically discovered categories by the tool so you get a fresh and broad overview of the problem space but then at the end of the day what you want is a complete reading list which papers will actually have to be so this is where the focus tool comes in handy and it enables you to very efficiently narrow down from a huge list of potentially relevant papers to an actual highly relevant reading list and this you can do without having to read the personal paper and using our tools you can cut down the time that is required for such a systematic mapping study up to 90% from several weeks to a few days actually and we've actually demonstrated that this works so at the end of the day you have your list of highly relevant papers so I mean this all very great and we're very successful with these tools we've named one of the top 10 innovative AI companies worldwide in 2017 by Fast Company we've raised over 2 million of seed venture capital we have many high-profile Gucci Coverage and top world university clients we have over 10,000 registered users who are free in premium tools and also AI trainers that help to improve our AI tools which are based on a proven and peer-reviewed technology that is also a large part developed in-house so yeah, that's all very great but as you remember there are also these three things down here that we wanted to fix initially also but we at Iris we believe that these problems are kind of on a different level so we think we can't want to set out to fix them on our own but rather we think this should be done in a community effort namely a grassroots movement of scientists, researchers, army departments universities students, librarians and innovators so we believe that a combination of AI and blockchain technology is the technological framework that is well suited for this change but at the heart of it all should be the community of scientists the people that actually produce the science so this is actually what we are proposing with Project AI that we will launch in the month or so so the goal is to build a community a decentralized community around building an AI tool that we call Knowledge Foundation Engine and this is set out to kind of fix the reproducible issues by sort of semi-automating the quality review so it should assist a peer reviewer during the review process but not only that but this is how you can think of it so it would be a community effort to build this tool to help with meditating knowledge and what we have in hand with it is also repository of already meditated knowledge which will be continually growing and of course it's also decentralized and accessible to everyone and by introducing a new token also for this community you can build new incentive models for rewarding people for things that have been doing for free jobs up until now like for example here in the papers we've heard many times that this is a serious issue but what I would like to think this token about is actually it's kind of a token of appreciation of the community for active community participation so you're incentivized to go ahead and actively participate in the community this is a thing if you're a scientist you want this thing to happen, you want science to change you should actively participate in your community and this is the incentive model that gives you actually also the reward for your contributions so why do we need blockchain why do we need a new token for this well of course I mean this is what it's all about we want this thing to be the tool for meditated knowledge and the connected repository of meditated knowledge to be not a property of a few entities that control this but this should be the property of the entire community or essentially human kind of course of course all members own it the one of the one of the defining features of blockchain of so every community member is basically on equal on equal footing so every member has equal opportunity but at the same time also faces equal scrutiny so contributions by members should not be should purely be judged on the quality of the contributions and not by their name or their presentation and of course by introducing a new token you can now start and design a new economic model from scratch through smart contracts and an automated central institution that will take action to actually balance the values of these different contributions that you can put into the community and services that you can take out of the community based on supply and demand and self-controlling process but still with this institution taking measures such that everything is balanced also against the outside so how can you now earn or spend tokens there are several ways that you envision how you can earn and spend tokens for example you could be an AI trainer and annotate data sets for the AI tools you could directly contribute codes to the knowledge validation engine you could find and retract and fix bugs or as a researcher you could directly publish your research straight into our validated repository and also peer review other people's contribution where the knowledge validation engine will be a great example of assistance to you and it will also help in further building and improving the knowledge validation engine on the other hand you can spend tokens for example by directly tapping into the AI engine so you can take a document and just have it validated I will explain this there you can also if you're an external company you can view this project actually as a community run open source project and you can just go ahead and build third party applications on top of it where you would gain access to the engine by paying for API calls so I mean it is basically open access but you pay a small fee for making use of it and connected with this you could then also use these tokens to directly pay for services that these third party services would supply for you which is actually also how it will work with our current and future premium tools at Iris so you can use these tokens to pay for our services actually to a much discounted price as compared to if you have a conventional subscription so the things that I've marked in green here are the ones that we envision to be the most imminent to be implemented on this of the community is this so the central goal of the community should be to build this knowledge validation engine which is an AI tool that can be loosely thought of semi-automating or assisting in the peer review process and the input would be a scientific argument and the engine would then extract and contain hypothesis and arguments will kind of link them against each other and validate them against all existing knowledge so it would kind of build a truth tree of the document and see what underlying assumptions or restrictions are not visible I guess so this is something like the argument map that David mentioned earlier so this is something that the ambition so in each of these underlying assumptions would be validated for it it would be shown or it would get a score how much this assumption or how much it relies on this assumption how much you can trust the hypothesis containing this paper and if there are any inconsistencies between arguments and the last so of course I mean this is kind of a if you have to go into more detail and of course it's a very big project but it's actually more realistic than you would think from an AI standpoint and the time that we envision I mean of course we envision for this would be to build this in about four years in the community effort of course subject to several milestones of service releases and finally the repository that would go hand in hand with it is of course like many of you also suggested this I mean we definitely have to not only publish polished oversold positive results there needs to be space for failed results or for reproducing science I mean it doesn't make sense if we keep on making the same mistakes over and over again because the guy next door doesn't tell you that it doesn't work so we also we also we embrace parallel publishing we do not care if you put your paper on our repository we don't care where else you put it you can put it anywhere if you put it on our repository it means it has the added value of the knowledge foundation engine and what I forgot to mention here is what I get asked quite a few times is it doesn't output a yes or no answer it's not like it's validated or it's not validated it's sort of like you get an automated report much like an well an assistant automatic report about the document with several scores and you can use it as help for manual review process so it's not a black and white thing it has many great shades between and many dimensions so it will be of course of an access decentralized it will be peer reviewed it also features ongoing review which means once it doesn't make sense that you review a paper once and then it's set in stone and you can't change anything anymore at any time go ahead and challenge review of your paper or you could write a follow up paper to show that some other paper is actually wrong and this would be in interaction of course it would be very nice to actually talk about this and see if we can make some integration there of course it's integrated and has natural collaborations tool third party tools that can be built on top of this IR engine so on top of the technological capabilities that watching offers of course as we've also heard today there's a lot of possibilities for new decentralized governance structures and at this point I want to mention that we as iris we want to be the initiator of this community we do not want to own this community so you can think of us as something like a service contractor or an elite developer for a certain initial phase to get the community up and running but after it's running for about one vision about 18 months or so we will actually step back and become one of the general members of the community so if we burn or give away our token such a default below 1% ownership cap that we will introduce such a no single member it can become all powerful and interesting to talk about this we will have a token sale like in May in early June where 75% of the raise fund will directly go into the community in form of escrow which will be released upon reaching certain milestones on the roadmap and of course we also all decisions that have to be taken in the community let's say about the roadmap or about questionable contributions or about challenges or something all of these will be centralized and democratized so it will be community decisions that every member can participate in based on the voting power which is based on the contributions how active they are in the community like the contributions they have added to the community also yes there are token states and also how long they've been part and active part of the community so like three months life right? yes so can you make a quick buck with this well actually with this bureaucrats proof of human work it's actually a perfect example of functional token and it's therefore far from an instrument of her term financial speculation it's much rather meant for natural holders that really believe in the added value of this project to themselves or to humanity so having said that it is very well possible in their potential for the token to indirectly grow in value since the services that you will get for the token will increase in quality over time and yourself as a member will be incentivized to add to the quality of these tools because this will add to the value of the token so the technology will improve the repository will naturally grow and also through third party tools that other companies do at the top there will be external corporate money funneled into the ecosystem so to wrap it all up I mean why again should we do it that way as a company well we need the knowledge validation engine we think we can't build it alone and we don't want to build it alone so this is why we're releasing this to the community but at the same time we will be at least in the first phase the main contributor to this engine so we will get steady revenues for the community at least as long as the community thinks that our contributions are useful and even if there were some other companies that would say okay well that's pretty cool let's use this engine and build something similar to IRIS we will have first move to advantage so we have been thinking about these things or actually the co-founder stuff for many many years and we know exactly where we want to go with so we will have first move to advantage but most importantly I mean if you remember we wanted to fix these five problems and many more so it's it's most important about impact and you know making science a fun and open environment and you want to do this in a community effort together so let's science the shit out of this together we do two questions okay then quick questions before the last question comment thank you for this very nice talk I would like to say that I don't see the ongoing period you're working because you hardly find the ongoing period to review your work and you cannot incentivize a hundred of period how would you incentivize it here you get tokens for it but the thing is it's a functional token that you yourself can then again spend in the community to have your work reviewed or validated we can continue tomorrow with this discussion okay and then like one more you quick like very interesting talk when you look at the discussion and the last 10-15 years around computable research you have this impression that okay set aside humanities for a moment yes but for science and technologies the research paper is really really a bad container to package knowledge and review with your approach acknowledges in a way yes you can take those nasty papers and break down what is actually in them and now my question is is this really a sustainable best service that you can deliver with your artificial intelligence background to the community because if you would somehow incentivize and help researchers to come up with structured knowledge from the ghetto why are they producing this this would be much more worthwhile because we wouldn't come up with all these crowded papers but instead have nice nano-publications clear structured statements within a paper whatever and one additional question if you have a look at AI research there's a lot of going on with this branch of explainability of what a neural network comes up with and in this area this would be clearly a mark of openness and open values perhaps it's in place so that I access added value from the machine but somehow it is also explained to me how that neural network came up with that platform absolutely I mean I wholeheartedly agree with you that we should not only consider the scientific paper as the only contribution you can make as a scientist