TOKEN2049 - Opportunities in The Web3 Data Economy - Trent McConaghy (Ocean Protocol)

hi everyone let's here I'm gonna go over here it's really a pleasure to be here today and to share with you what we're kind of at ocean as well as opportunities that you might find interesting with respect to this emerging data economy let's get started so back in 2008 you all probably remember there was a financial crisis a lot of bad things happened for example there was lines of people outside banks because the banks are shutting down and they wanted to get their money but of course not your keys not your money so they didn't get their money in very many cases that led directly to the creation of Bitcoin which sparked the opening up of money so we went from this previous economy of money a shadow money economy controlled by the Fed controlled by the banks and now this token economy that has been emerging is really about being transparent and permissionless so the token economy is opening up money and it's led to an explosion of opportunity this very conference here is a wonderful example the trade show out there two ballrooms worth of both of keynotes and talks and so on and you know billions and billions of dollars worth of value created and we're still just the very very beginning and in this economy this token economy you can view it as really three levels there's the base layer where we have a reserve currency Bitcoin people will argue is that's a bit the reserve currency of the token economy there's an app platform with the unit of exchange aetherium and several others have emerged as well as a funding platform for things like IT O's and this sort of thing once again a theorem and many things on top of that we have the utility last mile this is the apps and as well on top we have the economic last mile and once again you'll see many of those boots out there with things like custodians and wallets whether it's in the cloud things like coinbase the hardware wallets or your software wallets a tracer and meta mask respectively we've got the exchanges and we've got the mining so all these are elements of this emerging token economy this isn't meant to be a complete picture but it kind of gives you a feel for what this looks like this is this relates later as you will see so so that's the token economy it's opened up money now let's talk about data I was at the UN last December and stood on stage and gave a talk about data and right before me I was pretty happy because a woman who is one of the leaders at the World Bank stood up and talked about how the digital economy is a data-driven economy and she gave some quotes in 2016 the global digital economy it's worth more than eleven trillion dollars it's huge right and it's growing so much that it will be one quarter of the world economy in less than a decade so we all know that the digital economy is huge here's the kicker and it was really cool to see that the World Bank themselves was saying this data is the fuel of the digital economy the digital economy this 11 trillion dollar industry is being driven by data a year ago The Economist published a headline article about data data as oil data as the world's most valuable resource many of the world's largest companies four of the top ten are actually based on data we've got Amazon Google Facebook Oberon and alike and the thing is it's a pretty closed economy right right now Facebook is buying their data from others and keeping it for themselves same thing with Google etc etc etc it's a closed economy it's opaque and that power is concentrated on a small number of organizations and individuals but we have an inspiration the token economy the token economy opened up money can we do the same thing for data can we move from this shadow data economy where they that is opaque with concentrated power and move to a web three-date economy and a transparent permissionless data economy while retaining privacy of course that's the question we've been asking and we might also ask what might it look like and today I'm going to talk to you a bit about how it has a base layer and economic layer of reserve currency the like it has a utility last-mile data science tools in the like and finally an economic last mile which is really surprising and interesting and I'm going to explore that here and of course in the state economy the economy opportunities are going to explode just like in 2008-2009 we got Bitcoin well now we have the emergence of a fundamental base layer for the data economy and we see that on top of that we will see many many opportunities with respect to custodians for data exchanges for data mining around data and other services utility to last mile and so on and so forth and so we're very excited about this and this talk I'm going to be elaborating on that so the first question is if we're gonna have a data economy we need to unlock the data so how do we go about doing this and this is the very very first question we started asking when we are working on ocean how do we unlock the data the demand side is easy ai and many other techniques really want the data so there's high demand it's just very hard to unlock how do you do it so there's two ways one way is connect the data halves with the data ones now inside every fortune 500 company and many other enterprises etc they all have tons of data but for most of them it's just sitting there latent it's gathering dust people inside they kind of have this feeling that their data is valuable but they're really really scared of letting it go of of losing control of getting fine because of gdpr and other privacy laws and so on at the same time you've got a whole bunch of AI experts that are in AI startups or in universities and so on and they're starving for data they really want data or even just basic analytics and there's very very very little overlap between these people who have the data and the people who want the data you know what the overlap is it's the companies that have created tremendous value by using AI to extract value from data Google Facebook and the like so the question we've been asking is can we merge these can can we connect the data haves the enterprises with all this latent data with all these AI folks who know how to extract value from the data and we believe that connecting is for the connective substrate a public utility network that no one owns or controls is the path so let's flash this out of what this can look like the goal is to reduce the friction to connect these data house with it data have-nots and that is via data marketplaces so on the Left we've got a whole economy of data marketplaces not one data marketplace but a bunch if you have just a data a single data marketplace centralized or decentralized it's gonna be a data silo so you need a whole bunch of these and then the enterprises and other organizations with lots of data are feeding into this you know government open data initiatives NGOs with dreams reams and reams of data on the other side you have the data scientists and the analysts with their Excel spreadsheets and so on and they're actually using their tools they're using Excel using data science tools like scikit-learn and so on and underneath you have a connective substrate that makes it really easy to build the data marketplaces and then that feeds directly into the data science tools into things like Jupiter notebook and other interactive tools so that's the goal is to reduce the friction to create these marketplaces to make it really easy where they have massive liquidity of data at the bottom a supply of data for everyone what's really cool once you have that then you have another benefit that comes for free you can do internal data sharing with provenance so this substrate at the heart of it it's really about decentralized access control and decentralize orchestration to be a bit more general but then what you can have is you have internal data sharing so let's say you've got a big multinational corporation with one branch here in Hong Kong and other branch in Berlin another branch in New York and you want to share the data that can be very very hard and you can also run into data privacy issues you know the data in Berlin can't leave German soil if it has any personal information on it for example so how do you actually make that accessible to share the data and this connective substrate is actually what makes that possible combined with the second thing I'll talk about so the second thing is you unlock the private data while preserving privacy so let's talk about private data the most valuable data is the private data it's sitting there behind the tiles people don't want to release it precisely because they know it's valuable so private and valuable it go hand in hand the thing is let's say you're a bank right and you wanted to predict the stocks or to do trading to do better trading or an insurance company and ideally you could somehow get all the data of your competitors even to build a really really great model to do prediction or maybe for credit card fraud detection or otherwise right so there's the data that you have but you really want you know 100 X or a thousand X that data that would be the ideal now here's the problem privacy right so you can't do this this is an absolute no-no it's show-stopping thanks to things like gdpr and so on and gdpr is a really great thing especially for the consumer I stand fully behind it from a business perspective if you want to extract value from AI that's the challenge so more data means both good things and bad things good things because you have more accurate models which mean you can have better business value you know sell more ads have a lower fraud etc but it also means that privacy is going down and the the exposure servers for exploits is going up right and we have all these hacks of things like you Equifax for 300 million credit cards and so on and so forth so you know every single big entity in the sense in a sense is a giant data honeypot no one like that so here's the solution bring the computer to the data you can unlock the value of the data while preserving privacy so you what you have is you have a decentralized layer in the middle ocean which does orchestration of these various computing steps you've got the private data behind a firewall you privately train your model whether it's a neural network or something else with a modeling algorithm supplied by the user you store the private model there and then you ask the model for predictions and only the predictions go up to the final user so here we're extracting the value from the data itself without exposing any of the personally identifiable information so we're actually resolving this fundamental paradox we we can have way more data for accuracy and business value but at the same time preserving the privacy this is a really big step so this is what ocean is about is about unlocking the data to unleash this open transparent data economy that preserves privacy as we've been working on ocean in the last couple of years we've been starting to engage with various partners on use cases so you can say this is what the beginning of a economy it looks like let's go through a few one vertical is automotive so one challenge in there is is autonomous driving we know that with autonomous driving we can eliminate the 1.2 million deaths per year from car accidents if these cars get way more accurate than humans which they promise to the challenges they need five hundred billion miles driven Toyota did an analysis and they realized that it would take them 20 years to get that much data and they realized they needed to share their data among Toyota GM BMW Ford etc etc etc so they formed an alliance called mobi and mobi is actually representative of 70% of the world's auto production we've been working with them since actually about a year before their inception and the CEO of mobi Chris ballinger has been a collaborator and to quote him he's proud to support us and our collaborators towards autonomous driving so this is one use case that we're very very proud of and working very closely towards another one is health so as an example a friend of mine he's a researcher in genetic programming he's building AI models for cancer he tells me he's a happy man if he built a model off of a data set of a hundred people this is genetic data so there's thousands of variables what that means the model is terrible you can hardly predict anything but imagine instead of predict making a model off of 100 data points you make a model off of 100 million now normally that would be a giant did a honeypot and it would be a privacy nono but with ocean you can actually resolve the privacy issue by learning across the different silos at once so in this particular case we are working with an organization called connected life and they are collaborating with TU Munich and other universities for chronic diseases they're focusing on Parkinson's Parkinson's to start and then after that other diseases as well the various forms of cancer Alzheimer's and so on and they're doing two things they have a trustless sharing environment between health data providers and consumers as well as a data Commons marketplace another vertical is the government of Singapore and in particular they're AI Singapore Department so we are working with them on research collaboration to help train up their next generation of engineers for decentralized AI and data sharing building data models in a privacy in fashion and so this is very useful strategically for governments because to quote Alexander Wisner grows AI advances six times faster when data is available so if you can take this insight and apply on a national scale where you've actually got great access to a nation by having more data then it's actually an excellent AI strategy at national levels we're also working with the government of Singapore's data authorities called IMD a and this is really about sandboxes for a regulatory fast lane so many Fortune 500 companies are headquartered in Singapore and so within those headquarters they're going to be running sand boxes where they don't have to have every regulatory solution up front they can run sand boxes for several months or even a year and once everything is fine from a privacy perspective and so on they can then they can unleash to the broader world and this allows them to iterate with new ideas and new technologies very quickly another example is in the world of retail their next billion and Unilever are working with us and the idea is that there's these role shops like the one pictured here in rural India and they have they sell say a thousand dollars worth of goods every month the Unilever's of the world want to be able to know what to sell these shops and they have no insight there's no Nielsen ratings for this so next billion is supplying applications for the people running these shops to say what what they're selling all the time supply that to Unilever Unilever is paying for this as data which is very valuable and at the end of the day the margins for the shop owners go up as much as 4x which is actually quite remarkable another example is agriculture so Syngenta is working to help small farmers manage their plots so how do you allocate your fertilizer better you know more in some places less than other places rather than uniform across the board and this in turn also helps to protect biodiversity and we have many other collaborators - Aviva Roche Johnson & Johnson St micro and more we're also working with other governments and other foundations such as AI XPrize AI frigging foundation grow Asia EXO and more so going back you know we now you've seen what this data economy is starting to look like we're just at the very beginning of this open transparent permission the state economy and you might ask yourself okay is there something in it for me what can I do and we're really at the very beginning this is 2008 2009 2010 in token land that's where we're at for data land right now so if you recall I've talked about how there is the the token economy and we've seen this explosion and all this opportunity you know billions and billions of wealth created now right now this we've got this new data economy that is just emerging and I've given a sampling of some use cases so at the base layer we've got ocean and it actually the way that ocean is designed it plays the role of store value as well as unit of exchange as well as apps platform one level up this is where we have the data science tools and so already ocean is integration into Jupiter which is the de facto standard interface that data scientists around the world are starting to use and Jupiter itself ties into things like tensor flow and scikit-learn all these things for things like neural networks and the like and we envision integrations with many many other tools to leveraging you know at the core level things like Python tends to flow directly XO on now here's the interesting part the last financial model this is unbounded opportunity just like we see these tradeshow booths out here for the token economy imagine five years from now we have a conference like this where there is a conference filled with tradeshow booths around data data custodians data wallets data exchanges and generalized mining for data so here let's talk about generalized my name you know mining in Bitcoin and so on is really about building these Asics the bit mains of the world and at generating money there with a notion this is actually more about data services so it's making data available storing data as well as compute so as examples from the web free world are things like polka dot at web through foundation file coin storage s3 as well as the compute side and nygma for privacy-preserving compute true--but we had a talk earlier today from single internet and so on and finally we actually have the web to world for example Azure and and Amazon so this is really generalized mining and there's going to be a lot of opportunities just in that but these other bullets are fascinating too so now if you think of what crypto wallets you've got a my ether wallet or you've got your trends are now you can use this to store and secure data right so before we talk about data security where you hope that your CIO or your assisted min is taking care of it right many many levels down on the stack no it's not your keys not your data right so whoever has the keys controls the data and and so there's way more tokens before and it's basically data security meeting token security we're going to have hardware wallets we're going to have paper wallets we're going to have custodians all of this just for data this trillion dollar economy that data will be also in the world of data exchanges so you might ask yourself we have had token exchanges we have the blind answers of the world the GD axes and so on well imagine if they start to adapt to their technology to data exchanges and with data there's tens of thousands millions of data sets so right now we might think we have a long tail of tokens for blockchain projects you know a thousand two thousand tokens well we're gonna be ending up with a hundred thousand or a million tokens one token for data set in the world of data and there's going to be a long tail it's going to be very exciting so there's gonna be way more tokens to sell and but it's not going to be quite specifically the same thing as data as token exchanges some of those data sets will be fungible some will be non fungible some might even have better payments with things like royalties in the like so we're gonna see all sorts of variety just around data itself and we can keep going within aetherium in other communities we're starting to see this explosion of defy with things like market makers you know swap or stable coins like maker Dow well now we can have data economy market makers we can have stable coins that are backed by data you know if we've got a trillion dollars worth of data assets why not make a stable coin backed in that if the future digital economy is a data economy why not make the stable coin back for that and also you know we have the Dharma's of the world which is about tokenized debt well we can have data debts we can have data insurance everything that we see emerging for the token economy will mirror into the daily economy how's that for exciting and of course some people are starting to see the opportunities so for example DCI digital commerce intelligence as well as connected life and others they're starting to build data marketplaces on top of ocean you know ocean is out there you can go to docks toad Ocean protocol comm and start building data marketplaces data Commons integrating to your favorite data science tools you can do that today the beta is coming up in two weeks production to follow shortly after so to conclude the token economy open money we all we have to do is look behind us and see the web three-date economy opens data while retaining privacy and with that we get an unbounded explosion of opportunities with data custodians data wallets data exchanges generalized mining and more how is that for exciting thank you