How Web3 Can Rebuild Truth on the Internet

Transcript

And we're just taking 2 seconds. Hi. My name's Yaniv. I founded The Graph back in 2017. Been working on Geo the last 4 years. And uh I'm going to be talking about redefining truth on the internet. So, I want to start with a question. Do you believe in objective truth? And what when I was growing up, I feel like this was a really well-understood concept. And it feels like in some ways it's gone a little bit out of fashion.

Uh you know, people are moving a little bit more towards this, you know, there's your truth and there's my truth. And uh you know, I can't help but wonder if, you know, this lack of belief in objective truth uh is a little bit responsible for the sort of nihilism that we're seeing in the world today. Right? There's a little bit of this, you know, if there is no truth, then and and nothing matters. Um And uh you know, if you don't have truth, then all you have is tribes.

Right? It's your tribe or my tribe. Right? There's no higher purpose that we're trying to aspire to or thing that we're trying to understand. Uh it really just becomes a battle. And uh and every side just wants to win. And uh you know, it really causes a lot of conflict in society. You know, I would love to be able to have, you know, conversations with my friends, with my family, um and not have it feel like it's this existential, you know, fight is polarization that we're seeing.

And this stuff really affects us. You know, in many cities around the world, you see, you know, rising homelessness, you see difficulty getting basic things done, you know, uh whether it's, you know, just building, you know, construction, creating housing, you know, any of these sorts of things. And I think so much of it is because we struggle to have quality conversations. We have, you know, policies that are made um that are not the best policies. And uh and it's easy to, you know, go down the wrong path. Uh but, you know, these things they they really matter.

And uh and it really affects our lives at a global and and and a local level. So, if you you ask yourself, you know, if you ask somebody in, you know, the the 1100s or something, like, what is truth? You know, they would say it's whatever's in the Bible, right? The word of God. And uh you know, we had the church, you know, working in tandem with uh maybe a monarchy. Um and uh and and, you know, the the word of God was the truth. And for, you know, many, many centuries, this would have been the answer to that question.

You know, then we had the scientific revolution. And you know, there we created a system. People wanted to say, "Hey, you know, if you want to try to assert that something is true, uh you're going to make a hypothesis, and we're going to test it, and we're going to use replication." And uh and I really love the the scientific method. Um and one of the things I like about science is it's a very humbling discipline. Right? You you almost start out with the assumption that you're probably wrong. And if you want to try to assert a fact, uh you have to rigorously test it. And most of the time, you'll be wrong. Uh but every once in a while, you might find a new discovery that pushes all of humanity forward.

Um you know, we have different types of institutions that have built up around providing the truth. And uh for for many decades, we had a you know, really strong press, um centuries even. Um and uh they built systems around, you know, being able to uh build trust uh with an audience using, you know, processes like, you know, verifying multiple sources, you know, to to corroborate, you know, evidence. And And uh and we had these, you know, newspapers that had built up trust.

Um but at the beginning, when the printing press was first released in uh in the 1400s, uh they actually had a crisis that was very similar to what we have today. Uh suddenly, anybody could print something. Uh and uh and and there were pamphlets that you would have around in different towns and cities. And it was really hard for people to understand what is true. You know, if anyone can print anything, and it's not just the word of God anymore, how do you know what you can trust? And they had the same exact problem that we have today. Uh but they did build systems um to solve this problem. And and uh and it was really important for uh for the evolution of society.

So, today we have the internet. And uh you know, as this was starting, you know, if you wanted to get an answer to a question, maybe you type it into Google. You know, maybe you'd Google yourself. You want to you want to find facts about somebody. And uh and maybe at some point this became seen uh as as being a a source of truth. And now, of course, today with LLMs, people are using Chat GPT, you know, Anthropic. And uh and you're asking these LLMs a lot of questions. And whatever they return becomes a sort of truth, you know? It's okay, according to Chat GPT, this is the answer.

Um but you have to ask, whenever you get a result, okay, but according to who? Is it something that was trained inside of the model that a bunch of people in an AI lab decided should be part of the training set? Is it, you know, web tools that's hitting Google, um and or, you know, whatever search, and that's how you're getting the response? So, you know, you have this tremendous concentration of power in these AI labs, um but we still don't have, you know, an open process that you can really understand um to answer this question, right?

According to who? Um So, uh you know, I have an idea for a system that we can use to answer this question. It's uh according to Elon. It'll be uh between those entities. So, the main

thing to understand here is most software applications that we use today use relational databases. Right? You have these tables, and they have a fixed structure. These are the fields, these are the relationships, and it all has to be known ahead of time. And so, that's actually not adequate for representing a complex world, where you're constantly discovering new relationships between things. And so, you need something that's more flexible. Um it's also uh is is really important if we want to have composability and interoperability across different types of applications. Right? If you have a single application, a relational database works well because you can you know exactly what you need for that app.

But if we want to enable open, decentralized systems and lots of different applications that can use the same, you know, knowledge base, then you actually need knowledge graphs. And this is a core part of the solution. The next thing you need is a way to organize that knowledge. And we do that through communities called spaces. So, each space can organize its own knowledge graph. And these spaces um are self-sovereign. They govern themselves, and they're arranged in the sort of subtopic hierarchy.

So, you start at a high level, and you can drill down into more and more detailed spaces. And each space, you know, can have a set of experts or people that, you know, understand that topic better that can organize the knowledge for that space. And importantly, we support pluralism. You know, this is some some of the problems that we have with these more centralized tools, like, say, Wikipedia, is that there's only one version of any given page. And so, you know, if there's a you know, the the the the COVID page on Wikipedia that says no, there's no chance that it was made in a lab.

Like that's it. That's the answer and there's nothing, you know, if you disagree with that, like that's the end of the discussion. Um so so we believe pluralism is really important because even though objective truth is important, nobody has a monopoly on truth and no one can claim that they absolutely know the truth and so you have to be able to support different perspectives. Um that's an important part of the process. Um but those perspectives should be competing with each other um and because they're part of the same system, they're actually part of the same knowledge graph, you can see where do these different communities disagree?

Um and that's really important uh and what is considered the global consensus at any point in time can change. Uh but you can have these different points of view that can coexist as part of the same system. And so these spaces, they have governance and uh and people can participate in the spaces, they can publish knowledge and uh and you have folks that can review that and think of this a little bit like a git, you know, for knowledge. You know, you can have uh conversations, review and you have like a history. So, you know, anything that you're you're querying and you're asking

in the knowledge graph, you can see how did it get into this state? Right? What was the conversation that was had and uh you know, who published it, who changed it? And so, I think having that kind of audit trail and um and provenance is really important for building trust. Um but only, you know, the the big sorts of changes really need to go through that sort of governance. Um you can also have things where it's very easy for any members to to publish things and and there you rely on things like rankings to help surface the best content. And so having, you know, generalized system for rankings is is is really important where any type of entity can have a score, you

know, relationships between things can have scores. Uh these things can get updated and this allows you to discover, you know, the best content and I think this is important both for questions of truth, but also for subjective matters. You know, I think uh you know, having good open information systems is really important for subjective things as well as objective things. Not everything is an objective fact, uh but it's actually very valuable to get different people's opinions and perspectives and you know, even things like, you know, art or restaurants or, you know, uh any type of thing, uh it's nice to have open information systems. People can share things, they can vote and you and we can have rankings and we can know according to

this community, these are the best, you know XYZ. Um we do think that in order to uh create a good system that converges towards truth, um that uh that that produces great quality, uh it's important to have reputation. So, you know, sometimes these crypto systems uh they're based on tokens, use token voting, things like that. Um you know, I think that it's very important actually that you can't bribe your way toward towards changing the truth. Uh just because you have more money uh doesn't mean that your opinions are more true.

Uh and uh you know, knowledge is not evenly distributed, right? Some people know more than others and you know, it's an interesting question, what are the factors that we would use to, you know, build into those uh weightings? Uh and it it's it probably a combination of lots of things. It could be, you know, your your work history, your education history, what have you produced? Um and I think predictions is really interesting. Now, you know, we have prediction markets and you know, if you've made a prediction on something that was, you know, controversial or, you know, not a lot of people believed in and then later turned out to be true, that's probably a a good indication that you're knowledgeable in that area.

Um you know, I think experts in different fields, they tend to be able to spot each other out. They're not always true, you know, that and and sometimes you can form into like cliques, but uh but but generally, experts tend to be able to find other people who are knowledgeable in a given area more than, you know, someone who who maybe isn't knowledgeable in that area. So, I think, you know, with all of these different types of things, um we can build up reputation and reputation is really important um I think toward for creating systems that converge towards truth.

And so ultimately, I I believe that our goal needs to be to build a meritocracy, right? An open system that anybody can participate in, but that rewards quality contributions, hard work, you know, people that uh uh that take the time to really understand things and and create value, you know, for others. And so, um you know, just like, you know, we do believe in objective truth, you know, we also believe that, you know, we need to be building a meritocracy and if we create a system that incentivizes people to do great work, um that's really important and good for society.

So, last year uh we released Geo Podcasts. It's an app that's built on the Geo knowledge graph and uh and what we've been doing is extracting from, you know, these really interesting conversations that are happening on podcasts every day uh where, you know, top experts go on podcasts, they have interesting conversations about the work that they're doing or things that are happening in the world and we extract claims and the topics of conversation and uh and a lot of knowledge that up until now has just been like embedded in and kind of sort of way in in in these podcasts. Um so we've been quietly building that up and uh and there's

actually a ton of just really interesting knowledge uh that that comes out of that. And today, uh I'm excited to announce Geo News, uh which is a new app that we're we're launching today um that's an aggregator uh from uh for for the top news stories that are coming out every day and uh and similarly, we extract the top claims um that are coming out of, you know, each of these news stories to start to create like a a historical record of what's happening in the world. And uh today, we're opening with AI, crypto, health and world affairs as being our first four spaces for crypto news. Uh we'll be expanding in the different uh spaces

that we support here. Uh we're going to be making it more interactive. Right now, we have a bunch of curators that are are part of our early network that are organizing this knowledge. Um but we want to open it up so people can can participate. Uh but it's a really nice reading experience and a big part of what we want to do is, you know, get out of this world of clickbait, you know, these algorithms that are trying to hack you and really provide something that is empowering and useful to you. Um that it's informative, uh that respects your time and uh and you know, we we believe that, you know, better knowledge is uh is empowering for people. Um so releasing this today, you can check it out at

news.geobrowser.io. Uh it's live now. And um you know, podcasts and news and uh we're going to be releasing more uh apps like this that are all built on top of the Geo knowledge graph. So, all of this knowledge is interconnected and um you can also use it in your own applications, you can use it from your agents um can access, you know, this this this news feed. Um and uh and we think that having all of this knowledge uh open, uh composable and also verifiable uh is is is going to be really useful.

Whenever we publish content here, it actually goes on the blockchain. So, you know, first, we serialize the knowledge in uh this new knowledge graph standard that we've created called GRC-20. We upload it to IPFS. Uh we publish it on chain and so there's a historical record of all of the changes and it's available to to everybody to use. And so, you know, a lot of people came into Web3 for the revolution. Right? We wanted to change how people cooperate and organize, uh to not have to depend on either, you know, the large banks or or tech platforms, uh to empower individuals and to give people freedom and

opportunity. And uh and I think this this vision is still very much alive and it's more important than ever. And I think we just needed the right tools before we could do this. Right? I think finance is really important. Open decentralized finance is really important, but so is knowledge and information. Right? There's you know, so many things can't be represented just as as money and in fact, you know, it's things that are happening in the world. And so with the right tools to actually organize, to coordinate, to share information and to better understand what's happening in the world and what we should do, right?

That's coordination, right? What What should our goals be? What should the policies be? How how how should we achieve these things together? Um I think we can make the change that we we all, you know, came here to make. So, I think we're closer than ever and uh we have to keep uh you know, the fire alive and and uh and now we have better tools to do this. So, if you're interested in participating, uh we just launched our curator program uh to bootstrap our initial network of curators. You can sign up here.

And we're going to be opening this up to our next cohort of curators very soon. And yeah, looking forward to to working with everybody to help create better systems of coordination to better understand what's happening in the world and to make the impact we came here to make. Thank you so much. >> [applause] >> Oh, yeah, of course.

See the next. Yes. Yes.