 OK, so, just to introduce myself, I am a cognitive scientist by training, so I began working in psychology and then moved towards the neuroscience side of things. And really the focus of my research over the last few years has been on neurodevelopmental conditions, things like autism and Down syndrome and language difficulties in kids dyslexia, that kind of thing. And I guess over the last few years I've become increasingly frustrated with the state of research and the way that we do research. And I guess that's partly that is sort of generic to the field of psychology, so I think psychology has really been the sort of ground zero for the replication crisis that is affecting science. So we now know that these problems are sort of fairly generic across the scientific spectrum, but it really began when people started realising that lots of psychology studies didn't replicate, even when we thought that they were pretty solid research. But there are also a number of problems that I think are more kind of germane to the kind of more specific research that I do in the sort of clinical side of cognitive research. So at the beginning of this year I co-founded Frankel Open Science, which is really a collaboration between researchers on the one hand like myself and people with more of a background in tech. My co-founder is Pete Godbolt, who's been working in blockchain and development for like four or five years. So that's, well, the blockchain side at least for four or five years, which is relatively speaking a long time, obviously. And so one of the things that we're really thinking about is trying to solve problems, and there's been lots of great discussions over the last couple of days. But one kind of comment that really struck out for me was something that Carmen said yesterday when what we're trying to do is build tools that make it easier for researchers to do good research, to do open research, to get the best value from the efforts that they're putting in. And, you know, what lots of people are trying to do is build tools, right? So building cool tools. But the strategy that we don't want to do is just kind of build tools and then just hope that scientists will come along and use them. And as Carmen mentioned, you know, different scientific fields have different needs. So really what we're trying to do is develop solutions to problems within this fairly narrow field, but then doing it in a way that then we hope sort of generalises to other fields that we can then sort of work with. So what are the problems that we're facing particularly in this sort of area of cognitive science with sort of clinical applications? So one of the sort of driving forces for me starting this idea was this idea that the tests that we're using to assess patients and people with different conditions in a cognitive way that it's a really sort of outdated model. Lot of it is done with pen and paper tests and they're really expensive. You can cost literally thousands of dollars to buy these tests and then you have to pay for the bits of paper that you're writing on each time you use it. It's really inefficient because it's pen and paper and then there's lots of opportunities for human error. And also it's kind of inaccessible for lots of different populations. So working with kids with autism, I found that that was lots of tests that we used, they performed really poorly but that's just because they find it difficult interacting with other people. But also accessible in terms of the fact that to administer these tests you need to be really highly trained. And so if you live in like a remote community or somewhere that doesn't have services of psychologists or speech therapists and so on, you just can't get assessed. So we wanted to develop applications that make all this easier. In the broadest game there's also the idea that if we were trying to make sense of conditions like autism we need large data sets and really most of the research that's been done today has been severely underpowered. And so we need collaboration between lots of different partners, lots of different researchers, people in clinical space. And if we're going to do that then we also need the infrastructure for sharing the data, for making sure that it's secure and it's all being done ethically. So really this is the kind of problems that we're trying to solve. So as Sonke mentioned we've just received funding from the Wellcome Trust for a small pilot application that we're building. This is working with a clinical neuropsychologist Professor Greg Savage who's an expert in dementia. So this first app is focusing on memory abilities in dementia although obviously in our diagram or animation this is using a child as an example. But the idea is that the first step of the application is that researchers collect data using the application. Nope, wrong way. And instead of the data sitting on the iPad or then being transferred to someone's hard drive it's automatically and securely archived in cloud storage initially. And that's done, it could be lots of different places, it could be somewhere like Fig Share or Open Science Framework. And that allows us to build into the data collection, this sort of process of data archiving which then makes it easier to share the data. And obviously that means that then the researcher has the ability to share that immediately with collaborators in the team. So none of this involves blockchain so far. But the blockchain comes in when we look at the payment for this. So instead of paying thousands of dollars you pay tokens, so you pay Frankel tokens each time you use these applications. And what that then means is that as you're collecting the data there's already a record of the fact that that data has been collected that goes immediately on the blockchain. So we're calling it metadata by metadata literally just meaning data about the data. So it would be pretty limited so just who it is that's collecting the data. Information about where the data has been stored but obviously if the data is still secure and unlocked you can't actually get the data itself. And there obviously wouldn't be any identifying data at that point. What it does mean though is that this metadata is immediately available to the public. So anyone can interrogate the blockchain, look at who's collected data with these applications so that there's a record of all the data that exists. And then the final stage would be at the point of publication or whatever point it makes sense to share the data more openly either publicly or to the scientific community. And that again would be situation specific. So the idea is at this point at any point the researcher can unlock the data. It makes it easy to share the data instead of having to sort of search around on the hard drive for the data. They can just change the access privileges to it. And then the final part is that as a reward for sharing your data openly you get a refund of some or potentially all of the tokens. So and initially the idea is that we are building these apps like we're doing with this welcome trust grant. But the idea is that the infrastructure that we're building around it we can then integrate with other applications that other researchers or other app developers build. And as I mentioned we're starting with this sort of fairly narrow focus on cognitive assessment tests. But we think that in the longer term this may be a model that might apply to other fields of research. So just to reiterate what's happened in this sort of scenario. So we've made it easy for researchers to share their data. It's allowing us to build standards for the data and metadata into the application rather than forcing researchers to kind of do this all independently. At MOSFest which I attended last weekend or the weekend before losing track of time. There was a big discussion among psychology researchers about developing standards that could be used across lots of different fields of research. And it's all fairly straightforward but it takes time and it's kind of not the kind of thing that researchers are going to do automatically. So we're making that all easy for them. We're providing a public record of the existence and location of the data so it makes it findable. And it's important that when we're thinking about the replicability problems that we're facing a lot of it is because of selective publication of data. So a lot of data is collected and then not published. So at the very least we're providing a record of all the data that has been collected not just the data that's been published. And one thing we're thinking about is including within that metadata on the blockchain the hash of the de-identified data. And then we can start thinking about how we can create a scientific supply chain from the data as it's being collected all the way eventually to what's coming out in the paper. And also we're adding incentivisation so there's incentive to share the data even if the study isn't published. So if you collect data and then you can't get the research published you still can essentially get your money back or your tokens back by sharing that data. And that might be useful to other people. And also there's an incentivisation to share your code or your applications that you're building as a researcher for collecting data and potentially any other analysis software. So you also get tokens in return. And importantly when we're talking about incentivisation these are tokens that their prime purpose is for collecting data. So we're not incentivising people with tokens that are exchanged for money, we're incentivising them with tokens that allow them to do more research. And we think if you're trying to incentivise scientists what you want to incentivise them with is the ability to collect more data and do more science. So just finally I want to talk about a concept that we talk about a lot at Frankel. This isn't mine, this comes from Pete, my co-founder and another guy he's working with called Tim. And they talk a lot about incremental decentralisation. So this is the idea that the end goal for a lot of what we're trying to do might be full decentralisation of research. And there's lots of cool ideas but it's going to be really hard to jump all the way there in one step. And so what we need to do is work with technologies like centralised technology that does already do a good job. So things like Figshare and OSF already exist and so it makes sense to work with those and eventually we might move towards things like IPFS. As the next step but we want to take it one step at a time. And the other part of it really is trying to remove the friction points. So if you want people to be using these we want them to make it easy for them to get into using cryptocurrency tokens. And that's not a straightforward thing to do. A few years ago Pete tried to persuade me to get into Bitcoin when Bitcoin wasn't worth what it is now. And I didn't because it took me so long to try and figure out how to create a crypto wallet. So one thing that we've worked on and that we now have and you can actually I'm going to talk about this tomorrow at the steps workshop. Is the Frankle Google powered wallet. So this is something that we think is pretty cool. So instead of having to go through all the rigmarole of creating a wallet you can literally sign up and create a wallet. And a theory of address that you control using your Google account. And it's quite clever how it works. I don't have time to explain it but I'll talk about it more tomorrow. And essentially it takes two minutes to click and then you have an Ethereum account and sorry an Ethereum address and a Google account. So if you want to know more come to the workshop tomorrow. And I guess that feels like about 10 minutes. Thank you very much John.