PI. Al Meets Blockchain: Exploring the Opportunities, and Challenges. P2. The Open Web is NEAR

[Music] [Applause] thank you hello everyone it's great to be here thank um KBW for inviting me and uh yeah I want to talk about where artificial intelligence meets blockchain and web 3 more broadly and how uh kind of we deal with opportunities and challenges so about me uh I'm actually originally from AI background I've worked uh on machine learning in AI for about 10 years I worked at Google research leading a team to understand natural language worked on tensorflow and I was one of the co-authors of a paper that has pretty much kickstarted the current uh kind of Trend in uh AI which is called attention is all you need and kind of introduced to technology called Transformers after that I have founded near which originally was an AI startup we were actually uh teaching machines to code and really trying to bring a noble way to interact with Computing um where everybody can become pretty much a software engineer and really program uh their Computing now it was a bit too early for this a five years early uh to be uh clear and uh one of the things that we end up doing is we build a crowdsourcing platform uh and we saw a need for a effective payment system uh that works on a global scale and this is where kind of our journey in web 3 started and we figured out that there's an opportunity to build something um in blockchain space and started near protocol in 2018 now near protocol uh has started with this broad Vision that we really want to bring Open Lab right ability for everyone to own their uh assets data and have power of governance and uh kind of since then been really bringing this Vision forward uh in a way way what we call now blockchain operating system that really propels and unites in much om new chain way as well open thatb but I'll let Marik to talk about that uh instead I will dive in a little bit closer into how AI is structured right now and how does this fit um the current Trend so one of the things that to know about AI is kind of this dichotomy of open and closed Source right uh there's in number of big companies that are working on so-called close Source uh AI although they publishing papers and kind of providing the uh some of the research they're doing uh you cannot run the results of this yourself you need to use their own servers the architectures are known right but the kind of weight the training that they've done is not and it takes an immense amount of compute to reproduce it now the reality is the open source like from my perspective always wins because because and this you know the reason why web 3 I think is eventually going to overcome web 2 is because the sheer amount of people and experimentation and incentives that's happening in open source is always dominating and what we see now with Facebook open sourcing llama and kind of really uh starting to match a lot of the capabilities openi has uh is proving that so what are the core pillars of you know any artificial intelligence uh system right there are three core pillars one of them is data so there's so-called large pre-training data this is what gbt stands for uh as well as so uh specific problem that you're trying to solve and you need data to find tune your model you have model architecture and weights that you kind of produce as a result of it and you have uh kind of amount of compute right gpus or accelerators that you need to train and run this model so these are three core pillars and they are uh interestingly intertwined with web 3 the other thing to know about AI right so this large language models language models is not a new thing it's been around for like 30 40 years uh but they've made you know huge major strides in the last five years as kind of this uh Transformers technology can uh has scaled up open source and kind of research Community is also really quickly evolving right so uh kind of the methods of reducing amount of compute needed reducing the way kind of and being able to do work on local machines is has extreme motivation right because there's a lot of people who have a single machine and not a lot of people who have access to Google clusters and finally uh there's a lot of kind of thin companies that have popped up which were like hey we going to use openai API and we're going to do our own own thing on top of it and kind of provide you a API this and you know they had a really Co cool tweets they had like you know tons of views but uh they have high churn because they don't really provide anything new and the underlying models that are continue innovating really disrup them right away we also have uh kind of an interesting change that happened right so before if there's some program on a computer right there's always a human that interprets the result there's a human who are entering the kind of inputs and interprets results and so for the first time this language-based models and is where Computing is now able to interact with Society directly right it's not any more the need for a human to be The Interpreter you have a direct kind of language output and which means it can interact not just with individual humans but also with kind of Society services it can interact with legal system with political system communication right it can run for an office uh potentially and this kind of creates a lot of Ripple effects in our society which we are definitely not ready for it also can have as any technology positive and negative effects right it has you know a lot of promise to you know help people and solve our problems to uh provide you know more personalized tutoring for children as well as you know better research assistance and the same time it can have have massive negative effects around you know misinformation and misuse in malicious ways and web 3 is really here to help structure the positive ways and mitigate the negative and I'll give you an example how it all works and I think the important part from my perspective is that the traditional regulations are not equipped to really um kind of contain this technology similarly what we've seen in web 3 where uh post the technology Evolution the amount of composability the amount of new things you can come up with right uh is just so much faster than the regulations can move and so we really need a more kind of software-based and practical approaches to really um kind of structure this so why now right why is this intersection as important now well AI has you know made amazing strides there's Hardware improvements over the past you know 5 years effective models uh kind of research as well as we have this chat GPT with a product Market fit uh of you know 100 million users we also have web3 now which has matured right we have now kind of scalable Technologies we have like effective transparency and reputation we have you know like effective markets been built around this that have enough on Rams and off-ramps and there's a lot of kind of common things that around like crowdsourcing and content management that are becoming more and more important uh in in intersection of both and governance is one of them so a very specific example right is because the society works with language uh the kind of all the interactions with financial institution with political uh bonent Etc now can be affected and they can be exploited by a malicious actor a human who can actually use these tools to at scale effect it and so the way to address is actually to use cryptography right notice I didn't say blockchain yet we need to use cryptography to authenticate a specific piece of content the traceability where data came from who who signed this now if you just do that on itself it doesn't actually give you the understanding uh who has done it because you know anybody can sign anything with something with something like you need a way to build a reputation around a specific identity and this is where blockchain comes in blockchain is an effective tool to build reputations and we now have social protocols actually to power that you know we have near social we have lens and a few others who are really powering that Community governance right we've been having experiments with Dows but it's really important now to have a diverse inputs around how do we kind of Judge and uh evaluate the models that they are actually safe and aligned with our values right we need to curate training data and evaluation data for these models in such a way that is not biased by single point of view or by by a small group of people and so we need a large community input to actually have this established we need to Define and kind of oversee agents uh and fund research so I'll talk about specific example because uh I think the AI web 3 and people need to work together in one kind of in one to solve really Global problems that humanity is facing so right now ai is seen from by some as a way kind of dystopian future right you know be that destroying Hollywood with generated content or uh kind of replacing everybody's job and the reality is AI is good at some things humans are good at others AI is really good at you know consuming large amounts of information being able to summarize it uh it's good at um pretty much kind of evaluating large amounts of uh possibilities and and you know figuring out what's the best given the specific criteria the blockchains are really good at coordinating uh people uh making sure everybody is voting creating transparency as well as creating fin like Financial system around that and people are good at Deep holistic expertise creativity and making decisions so let's look at a specific example where we combine all of this and to solve one of the kind of bigger challenges Humanity faces solving cancer uh this requires coordination of large number of people around the world it has massive amounts of budget uh that currently is not transparent how it's used and it kind of requires a ton of management across the whole industry so let's start a cure cancer Dow right this will be a dow of you know experts and institutions hospitals uh that come together to really um address this problem and so they will share this mission of curing cancer they specify kind of a small group of experts who will be running it and finally they'll initiate a dow where the CEO of the Dow the operator the manager the coordinator is an AI an AI agent and so this governance board sets up the mission the kpis around how we measuring success as well as you know procuring and helping with raising budget for this and uh the task for the AI agent now is to start recruiting the you know researchers and people research institutions to participate in it as well as to come up with a step-by-step plan and sub projects which are going to be uh moving Us in that direction as new institutions are joining such AI is on boarding them providing them context uh is able to summarize all the other work that has been done or in progress in the ecosystem and make sure that they are on their task and kind of in the specific Lane and not blocked by anybody else now the board this kind of the community of researchers that are established as Dow and potentially new ones who join are able to oversee and you know uh override if they think that AI is doing the wrong job it also interestingly and differently from people can provide a specific examples where AI is doing uh job wrong and and in a way have procured training data for this AI to uh adjust Behavior right so we can use the you know very effective way of pretty much coaching this AI right through training data to make different decisions or make different um kind of suggestions uh by procuring training data and giving examples right and you can use experts in management or you can use experts in uh specific domain expertise to really uh get this data you have this agent that is now able to update road maps keep everything up to date able to generate uh what is the up kind of the current state of the research provide you know uh like daily weekly monthly updates to all the stakeholders as well as very specific context to the people who are working and over time it can expand right it can even spawn new subd Dows which are also run by independent a AIS where they have a specific subtask for example curing a specific type of cancer where where you know maybe different type of research is needed it can coordinate the treasury and the funds of the uh of this Dow by pretty much being the ACT actor that creates proposals and you know they can be validated by the community and so the board is really there to oversee and kind of really work in the regular you know how we see boards and communities overseeing the uh execution of regular uh organizations but they are really have an active part now to be able to coach as well as they're able to uh participate in the discussions so this vision is not possible without bringing web 3 and AI together right you're not able to do things like this with just an AI running on centralized server because they're not able to um participate in the economy they're not able to pay people they're not able to uh do this you need the accountability and transparency of every action to be able to then uh not just audit but also give feedback and have uh right feedback loop the kind of streamlining that comes from this AI is really the way to start coordinating right in web 3 we had a very kind of crisis of dows that happened because although in the beginning people were really excited about coordinating Dows what happened over time is that it's a really hard work there's a lot of kind of people and all of them are stakeholders and all of them needs to be unboarded given context produced uh content and coordinated and it's a really hard job that people were actually not paid uh much to do and so leveraging the kind of the benefits that we already see from you know summarization and ability to kind of create content tailored for specific people we can now use this to kind of uh enhance coordination in this decentralized organiz ganizations and finally we can monitor and kind of create risk Frameworks around this because again everything is transparent you know exactly the models used you know and uh how to evaluate the risks of this and this is just a high level example right of how a very specific you know kind of use case can be put together using this uh Technologies and really starting to benefit something that's bigger than just you know AI research or web 3 kind of Our Own small pond but actually Mak an effect on the real world and I think this is what we all should be striving for actually getting out of just our somewhat you know uh sometimes circular economy and really starting to think how do we affect the change in the world we having around us so just to kind of mention some of the things that are already in progress right so uh I mentioned you know Three core pillars data being one of the core pillars and blockchain is extremely powerful play to create marketplaces and so data kind of crowdsourcing and Marketplace is one of the kind of biggest needs right now because although there's a lot of pre-training data that's available as specific problem uh sets right everything from you know how do you answer to this question to you know is this sentiment positive or negative is something that everybody's looking for and so creating a marketplace where everybody in the world can participate in doing tasks and anybody can actually procure data is extremely important and it also allows over time to have this onchain agents uh to actually participate in this economy as well now looking forward a little bit further we'll have more Specialists right so it's not just anybody participating but as you are doing work you're actually uh proving out that you have skills in specific domain and now you can have this Marketplace which are more um more specialist and the interesting thing here it doesn't matter if on the other side there's an AI or or a human doing the work right the marketplace on the kind of in blockchain actually removes that uh uh kind of you know Divergence and now creates kind of a singular you know future work where you can like hey I need this creative written and maybe it's written by AI or blockchain if you have a way to uh kind of qualify it or there other people verify the quality finally I mentioned kind of this uh onchain agent Dow coordination and I think this is one of the kind of big use cases that we'll see like right now will be more as a experiments but it as it mature so it will become more and more dominant as a way to coordinate and govern both blockchain projects but as well and more real world projects and so there's few kind of companies are starting to work on this finally there's a lot of opportunity right uh kind of moving away from the centralized uh services like open AI mid Journey ET Etc and actually moving the the generation on edge right be that on your phone on your device on your TV allowing actually hyper personalization right it can know exactly what you know videos you watched before or what Communications you have and create content and create videos and images just for you and so that requires you know smaller models on edge as well as a um kind of ways to potentially decentralize some of the computation you have fan owned IP right so this is more a web 3 part where the IP itself probably is going to be uh generated right so it's user content generation but then you have Brands and styles that are created by uh either individuals or teams or and then can be owned by different people and they receive the part of revenue from the generated content and so all of that really like is infrastructure to tie together all this desperate pieces uh and desperate kind um kind of um participants in the ecosystem finally the kind of decentralized science so-called research funding Beyond just you know investing in bunch of startups which potentially don't make bunch of Revenue but creating a really interesting technology there is a new way of venture where we funding research uh and really allowing to capture value from the open source technology right and blockchain and kind of tokenization is one of the ways to do that so uh this is just a c code to sign up for your tasks for the AI crowdsourcing and uh this is just an example of a project we're doing and uh yeah I'm going to pass this to Marik thank you thank you very much IIA um we're going to stay on topic but go and talk more about the open web in general so all the way from Switzerland the CEO of the near Foundation Marik flon is going to talk to us about um how the open web is near big round of applause thank you very [Music] much hello uh Korea great to be here and building on uh elia's perspective on AI and web 3 and how things interface what I want to talk about today is an open web why we need an open web and actually how we think about an open web and what we're doing about it at near when we look around the world today we're living through several crisis we are living through a climate crisis we are living through divisions of our society the left the right we also just have been through or is it even finished Corona virus and who knows IIA was mentioning you know having to treat some of the largest diseases in the world but who knows what's coming next and not only that but I don't know how you feel every day when you wake up in the morning and your phone is basically shouting at you with like thousands of messages in all directions just trying to grab for your attention the fundamental problems we actually have make it really sound like it's bad news everywhere and actually guess what the reality is that bad news is good for clicking because if you click it will generate more revenue and so we are constantly fueled into this machine of clicks and bad news and which also divides more much more fundamentally as a society one of the problem of the way the current web system is build is that actually the centralized web as we know it is all about profit it is not about people and what's happening with that is that it's actually causing breaks down and therefore all the crisis that we're having I'm of course exaggerating but the reality if you think of the machine and the cycle we currently are in that's what's happening that's why we need need an open web and so what is an open web well an open web fundamentally flips that definition of central parties and basically milking and using each and every one of you with your clicks for your data but actually giving that back to you it is actually having to build a web that is without the influence of large corporations or Gatekeepers and that gives everybody everywhere in the world an equal chance to participate and ownership of your data your ideas your efforts and that's what we are building at near that's what we believe in how can we build this open web a better web so what is near you heard from uh IIA our co-founder on the origin of near and near actually started as an AI project and because there was a fundamental problem in actually solving and having ability to actually tackle and have you know thousands of people around the world being thoroughly rewarded for the work that they were doing there was a need to create a better layer one and that's how near evolved then into a very powerful very simple scalable secure and sustainable uh layer one more recently we've continued our journey on the stack and the entirety of the technology that is being built across the near ecosystem and we've evolved into a blockchain operating system but our mission remains fundamentally the same it is about getting 1 billion users into this open web I mentioned that but at the corent and fundamental principle of near when we've been actually started was how can we have something that is so simple so simple for developers to be able to develop with whatever language they know so simple for users to actually use because that's how we break down real entry to barriers it's like having things that are extremely simple to enter into security scalability of course are fundamentals that are needed as you think of bringing uh billions of users on chain and last but not least if we are engaging on this journey of leveraging technology to build the next future well it better be a sustainable future a better future that thinks not only about sustainability in terms of ecology but also participation and inclusion near has evolved and continues evolving into what we call a blockchain operating system and so part of this blockchain operating system is actually all the different components that over time we've been able to actually see being built on there and for example fast ass is one way to log in seamlessly and be on boarded into near but also other uh layer three layer two and and layer one uh ecosystems now at near Foundation our mission fundamentally and first and foremost remains on building this open web and enabling an ecosystem of participants to be self- sustainable and over time reach the goal of what the open web can be and ultimately onboarding 1 billion users where are we at on our journey well so far 29 Mill million wallets have been created uh and we are at 2.1 monthly uh active account you can see on the right hand side that there is a huge stick coming up and I'll talk more about that now the next part of the presentation and what I want to talk about is actually well how we how do we do that how are we strategically thinking of building this open web and having all the pieces coming together the first thing I want to mention is actually long-term view is very important this is not about creating F fomo this is about building real active engage users that will truly participate in the vision that we have for this open web we spend a lot of time thinking about well if the value economic model for the web as we know it is broken then what is the value economic model for web 3 how does it work and the reality is that it is very complex it interfaces many different participants and many many different uh pillars but fundamentally what it does it that it Acres value back into open- Source technology and fundamentally also what it does that it actually incentivizes the participants and it rewards the participant of that economical model to be engaged in the model to truly get ownership of what they are creating or what actually they're engaging with if you want to learn more about that uh we actually published our Q2 transparency report and we went through actually the details of how we think about this model and how it actually helps inform our entire strategy so that's the first part like what's actually the model and how does incentive and the mechanic of the machine flows the second part is well it's great if you want to have a billion users but how are you actually going to do that and our fundamental belief is that this needs to be a two prong approach you need to actually bring Web Two into web 3 I think if anyone of us standing here think that we can build web 3 purely just from the ground up it's not going to happen the reality is that today you have billions of users who are in web web 2 and to get to billions of users in web 3 and in an open web well this will only happen if actually web 2 gets onto the journey and that's what we do with our strategic partnership and I'll showcase to to you some of the use cases that we've been doing but equally that's not enough because if we only leverage web two companies and try to onboard them into web 3 we are going to miss out on actually the truly grassroot opportunity and The Innovation that is creating things that we don't know yet about and so that's why our approach is too prong it's thinking about what is top down but also how can we Empower grassroot community so strategic partnership what do we do and how do we do that today we are actually uh recently over the last few days uh we started uh seeing coming online on mainnet that's the spike that you saw uh on my previous slide a platform that is called Kai Kai and Kai is actually a loyalty program based out of Singapore that had a fundamental business problem which was that for their loyalty program to actually be uh beneficial and to have a real viability in the business model if they were using systems such as tripe or existing Financial rails where the entire economic model was falling apart and so this is a business that has hundred thousands and million of users that had a true business needs to actually onboard its users into web 3 and really use the the rail of web 3 not only that but Kai is also soon launching a new product which is called K now which is something that you be able to find on your lock screen and creating therefore a loyalty program this is an example of a web two startup that started as a web2 that has strategic impediment and problems that can only be solved by web 3 and so for us in our view this is a perfect use case of actually on boarding onto web web three millions of users a second use case for which we are actually known and if you haven't downloaded the app yet I encourage you to do so uh it's sweat coin with sweat the idea is that you walk and you earn but not only that you earn and you get on boarded into web 3 and now you can actually truly participate into web 3 and start having a taste of well what is your effort actually really worth and actually how can you engage into other platforms and start you know making different experiences out of that sweat has over 120 million users around the world uh and is still looking at launching more and more markets and opening up under underneath leveraging the near uh platform play Amber and in particular when we talk about gaming is a very interesting game in gaming there is a lot of different possible angle of attacks for let's say web 3 and the reality is that gaming fundamentally transformed the web as we know it gaming is what push pushed forward mobile web and there is no reason not to believe that actually gaming is not going to be the next generation of what will actually enable us to create the next web 3 now the reality is we still haven't found what is that magic recipe and so our approach to gaming is multifold one is actually to Think Through casual games like play Amber who have millions of users how can casual games actually start onboarding millions of people into crypto but we also have other approaches with more traditional Web Two studio and also in games in thinking through well how can actually gaming help reinvent and the next phase and enable the next phase of web three cgp is a strategic partnership that we've done so um cgp is about sailing uh it is actually uh it's been built by Larry Ellison which is a founder of Oracle and the spirit of cell GP is well we live in a world that needs to be sustainable so maybe there is an alternative to Formula 1 and maybe that alternative can be on water and it can actually be powered by wind more even and more than that the teams that are on every boat are also diverse teams so it's not just about an only male team but it's actually about also diversity and having different representations cgp approached near to work on their next Generation loyalty program and how can we actually embark on a journey where fans truly own a team and so that's what we have today which is two partnership but also like a DA which owns a first a sports team which is actually a professional sports team we're in Korea and our Korean team is doing a phenomenal job at actually working with some of the largest players that are in the market and here I wanted to give two examples that we announced recently talking again about gaming one is marble X and the other one is uh Vortex gaming so that's our strategic partnership right thinking about how can we on board you know the Next Generation and the users that are in WID to thinking and leveraging actually the usability and the simplicity of NE technology but as I mentioned that is not enough and so one of the efforts that we consciously make is community how can we engage our community to continue building what near needs to have as a technology to enable the open web Vision how can we actually have the community to self-govern itself so that ultimately the foundation is not is not needed two efforts we're doing on that front one is a dow called near Dev on near Dev there is actually working groups that are continuously having uh open up discussions but also creating the next generation of components that are needed to evolve the near stack for developers out there I highly encourage you to have a look at uh that we've recently seen in our community something that is probably one of the largest social experiment into web 3 and that is called the near digital community and ultimately what this is doing is first and foremost we have a trust an instrument and a treasury that is self-governed by entirely the community and the community is organized around three different houses as I speak we are going through our elections and so all of that is on chain and basically we are starting to test the limit of a gigantic experiment to see how we can have the community truly owning funds and being able to disperse funds to jointly work together on achieving this vision of an open web again if you want to have a look at that I highly encourage you to look at a NE digital Collective the last part is what we realized in particular through last year is that it is not enough to actually just let the community or just to offer grants and and leave people and wish them good luck that just doesn't work and what we realize is actually there is a need for a platform and a place where people can truly exchange knowledge and can actually lean on each other and learn from each other and that's what we've started building with near Horizon so near Horizon uh was recently launched and today has over 242 project that are listed but also has over 89 contributors that are basically offering their time their skills and even sometimes funding we see that as a truly onchain accelerator and incubator and we participate with some of the we partner actually with some of the largest uh uh some of the largest incubators um out in the world to make that Vision come to life again here if you're a project or a builder highly encourage you to uh look at uh to look at that platform and that's what I have so the conclusion of my talk is that the current web in which we are doesn't work because fundamentally the economical model is broken and so there is an impediment and a need for creating an open web creating an open web cannot be done in isolation and recurs actually thoughtful strategy impediment and the right underlying technology and of course the open web is not going to be solved only by near we are a player within that ecosystem we're thrilled to participate in that but we highly encourage you to join us into the journey thank [Music] you