Episode #106: Building Apps on the Decentralized Web with Nader Dabit


June 21, 2021 • 59 minutes

On this episode, Jeremy chats with Nader Dabit about Edge & Node's graphQL API for querying blockchain data, how this and other decentralized protocols power the Web3 movement, what types of applications you can build with them, why you'd want to, and a whole lot more.

Watch this episode on YouTube:

About Nader Dabit

Nader Dabit is a web and mobile developer, author, and Developer Relations Engineer building the decentralized future at Edge and Node. Previously, he worked as a Developer Advocate at AWS Mobile working with projects like AWS AppSync and AWS Amplify. He is also the author and editor of React Native in Action and OpenGraphQL.

Nader Dabit Twitter: @dabit3
Edge and Node Twitter: @edgeandnode
Graph protocol Twitter: @graphprotocol
Edge and Node: edgeandnode.com
Everest: everest.link
YouTube: YouTube.com/naderdabit
What is Web3? The Decentralized Internet of the Future Explained


Watch this episode on YouTube: https://youtu.be/pSv_cCQyCPQ

This episode is sponsored by CBT Nuggets and Fauna.

Transcript

Jeremy: Hi everyone. I'm Jeremy Daly and this is Serverless Chats. Today I am joined again by Nader Dabit. Hey Nader, thanks for joining me.

Nader: Hey Jeremy. Thanks for having me.

Jeremy: You are now a developer relations engineer at Edge & Node. I would love it if you could tell the listeners a little bit about yourself. I think a lot of people probably know you already, but a little bit about your background and then what Edge & Node is.

Nader: Yeah, totally. My name is Nader Dabit like you mentioned, and I've been a developer for about, I guess, nine or ten years now. A lot of people might know me from my work with AWS, where I worked with the Amplify team with the front end web and mobile team, doing a lot of full stack stuff there as well as serverless. I've been working as a developer relations person, developer advocate, actually, leading the front end web and mobile team at AWS for a little over three years I was there. I was a manager for the last year and I became really, really interested in serverless while I was there. It led to me writing a book, which is Full Stack Serverless. It also just led me down the rabbit hole of managed services and philosophy and all this stuff.

It's been really, really cool to learn about everything in the space. Edge & Node is my next step, I would say, in doing work and what I consider maybe a serverless area, but it's an area that a lot of people might not associate with the traditional, I would say definition of serverless or the types of companies they often associate with serverless. But Edge & Node is a company that was spun off from a team that created a decentralized API protocol, which is called the Graph protocol. And the Graph protocol started being built in 2017. It was officially launched in a decentralized way at the end of 2020. Now we are currently finalizing that migration from a hosted service to a decentralized service actually this month.

A lot of really exciting things going on. We'll talk a lot about that and what all that means. But Edge & Node itself, we do support the Graph protocol, that's part of what we do, but we also build out decentralized applications ourselves. We have a couple of applications that we're building as engineers. We're also doing a lot of work within the Web3 ecosystem, which is known as the decentralized web ecosystem by investing in different people and companies and supporting different things and spreading awareness around some of the things that are going on here because it does have a lot to do with maybe the work that people are doing in the Web2 space, which would be the traditional webspace, the space that I was in before.

Jeremy: Right, right. Here I am. I follow you on Twitter. Love the videos that you do on your YouTube channel. You're like a shining example of what a really good developer relations dev advocate is. You just produce so much content, things like that, and you're doing all this stuff on serverless and I'm loving it. And then all of a sudden, I see you post this thing saying, hey, I'm leaving AWS Amplify. And you mentioned something about blockchain and I'm like, okay, wait a minute. What is this that Nader is now doing? Explain to me this, or maybe explain to me and hopefully the audience as well. What is the blockchain have to do with this decentralized applications or decentralized, I guess Web3?

Nader: Web3 as defined by definition, what you might see if you do some research, would be what a lot of people are talking about as the next evolution of the web as we know it. In a lot of these articles and stuff that people are trying to formalize ideas and stuff, the original web was the read-only web where we were not creators, the only creators were maybe the developers themselves. Early on, I might've gone and read a website and been able to only interact with the website by reading information. The current version that we're currently experiencing might be considered as Web2 where everyone's a creator. All of the interfaces, all of the applications that we interact with are built specifically for input. I can actually create a comment, I can upload a video, I can share stuff, and I can write to the web. And I can read.

And then the next evolution, a lot of people are categorizing, yes, is Web3. It's like taking a lot of the great things that we have today and maybe improving upon those. A lot of people and everyone kind of, this is just a really, a very old discussion around some of the trade-offs that we currently make in today's web around our data, around advertising, around the way a lot of business models are created for monetization. Essentially, they all come down to the manipulation of user data and different tricks and ways to steal people's data and use that essentially to create targeted advertising. Not only does this lead to a lot of times a negative experience. I just saw a tweet yesterday that resonated a lot with me that said, "YouTube is no longer a video platform, it's now an ad platform with videos in between." And that's the way I feel about YouTube. My kids ...

Jeremy: Totally.

Nader: ... I have kids that use YouTube and it's interesting to watch them because they know exactly what to do when the ads come up and exactly how to time it because they're used to, ads are just part of their experience. That's just what they're used to. And it's not just YouTube, it's every site that's out there, that's a social site, Instagram, LinkedIn. I think that that's not the original vision that people had, right, for the web. I don't think this was part of it. There have been a lot of people proposing solutions, but the core fundamental problem is how these applications are engineered, but also how the applications are paid for. How do these companies pay for developers to build. It's a really complex problem that, the simplest solution is just sell ads or maybe create something like a developer platform where you're charging a weekly or monthly or yearly or something like that.

I would say a lot of the ideas around Web3 are aiming to solve this exact problem. In order to do that you have to rethink how we build applications. You have to rethink how we store data. You have to rethink about how we think about identity as well, because again, how do you build an application that deals with user data without making it public in some way? Right? How do we deal with that? A lot of those problems are the things that people are thinking about and building ways to address those in this decentralized Web3 world. It became really fascinating to me when I started looking into it because I'm very passionate about what I'm doing. I really enjoy being a developer and going out and helping other people, but I always felt there was something missing because I'm sitting here and I love AWS still.

In fact, I would 100% go back and work there or any of these big companies, right? Because you can't really look at a company as, in my opinion, a black or white, good or bad thing, there's companies are doing good things and bad things at the same time. For instance, at AWS, I would meet a developer, teach them something at a workshop, a year later they would contact me and be like, hey, I got my first job or I created a business, or I landed my first client. So you're actually helping improve people's lives, at the same time you're reading these articles about Amazon in the news with some of the negative stuff going on. The way that I look at it is, I can't sit there and say any company is good or bad, but I felt a lot of the applications that people were building were also, at the end goal when you hear some of these VC discussions or people raising money, a lot of the end goal for some of the people I was working with were just selling advertising.

And I'm like, is this really what we're here to do? It doesn't feel fulfilling anymore when you start seeing that over and over and over. I think the really thing that fascinated me was that people are actually building applications that are monetized in a different way. And then I started diving into the infrastructure that enabled this and realized that there was a lot of similarities between serverless and how developers would deploy and build applications in this way. And it was the entry point to my rabbit hole.

Jeremy: I talked to you about this and I've been reading some of the stuff that you've been putting out and trying to educate myself on some of this. It seems very much so that show Silicon Valley on HBO, right? This decentralized web and things like that, but there's kind of, and totally correct me if I'm wrong here, but I feel there's two sides of this. You've got one side that is the blockchain, that I think some people are familiar with in the, I guess in the context of cryptocurrency, right? This is a very popular use of the blockchain because you have that redundancy and you have the agreement amongst multiple places, it's decentralized. And so you have that security there around that. But there's other uses for the blockchain as well.

Especially things like banking and real estate and some of those other use cases that I'd like to talk about. And then there's another side of it that is this decentralized piece. Is the decentralized piece of it like building apps? How is that related to the blockchain or are those two separate things?

Nader: Yeah, absolutely. I'm a big fan of Silicon Valley. Working in tech, it's almost like every single episode resonates with you if you've been in here long enough because you've been in one of those situations. The blockchain is part of the discussion. Crypto is part of the discussion, and those things never really interested me, to be honest. I was a speculator in crypto from 2015 until now. It's been fun, but I never really looked at crypto in any other way other than that. Blockchain had a really negative, I would say, association in my mind for a long time, I just never really saw any good things that people were doing with it. I just didn't do any research, maybe didn't understand what was going on.

When I started diving into it originally what really got me interested is the Graph protocol, which is one of the things that we work on at Edge & Node. I started actually understanding, why does this thing exist? Why is it there? That led me to understanding why it was there and the fact that 90% of dApps, decentralized apps in the Ethereum ecosystem are using it. And billions of queries, companies with billions of dollars in transactions are all using this stuff. I'm like, okay, this whole world exists, but why does it exist? I guess to give you an example, I guess we can talk about the Graph protocol. And there are a lot of other web, I would say Web3 or decentralized infrastructure protocols that are out there that are similar, but they all are doing similar things in the sense of how they're actually built and how they allow participation and stuff like that.

When you think of something like AWS, you think of, AWS has all of these different services. I want to build an app, I need storage. I need some type of authentication layer, maybe with Cognito, and then maybe I need someplace to execute some business logic. So maybe I'll spin up some serverless functions or create an EC2 instance, whatever. You have all these building blocks. Essentially what a lot of these decentralized protocols like the Graph are doing, are building out the same types of web infrastructure, but doing so in a decentralized way. Why does that even matter? Why is that important? Well, for instance, when you live, let's say for example in another country, I don't know, in South America and outside the United States, or even in the United States in the future, you never know. Let's say that you have some application and you've said something rude about maybe the president or something like that.

Let's say that for whatever reason, somebody hacks the server that you're dealing with or whatever, at the end of the day, there is a single point of failure, right? You have your data that's controlled by the cloud provider or the government can come in and they can have control over that. The idea around some of, pretty much all of the decentralized protocols is that they are built and distributed in a way that there is no single point of failure, but there's also no single point of control. That's important when you're living in areas that have to even worry about stuff like that. So maybe we don't have to worry about that as much here, but in other countries, they might.

Building something like a server is not a big deal, right? With AWS, but how would you build a server and make it available for anyone in the world to basically deploy and do so in a decentralized way? I think that's the problem that a lot of these protocols are trying to solve. For the Graph in particular, if you want to build an application using data that's stored on a blockchain. There's a lot of applications out there that are basically using the blockchain for mainly, right now it's for financial, transactional reasons because a lot of the transactions actually cost a lot of money. For instance, Uniswap is one of these applications. If you want to basically query data from a blockchain, it's not as easy as querying data from a traditional server or database.

For us we are used to using something like DynamoDB, or some type of SQL database, that's very optimized for queries. But on the blockchain, you're basically having these blocks that add up every time. You create a transaction, you save it. And then someone comes behind them and they save another transaction. Over time you build up this data that's aggregated over time. But let's say you want to hit that database with the, quote-unquote, database with a query and you want to retrieve data over time, or you want to have some type of filtering mechanism. You can't do that. You can't just query blockchains the way you can from a regular database. Similar to how a database basically indexes data and stores it and makes it efficient for retrieval, the Graph protocol basically does that, but for blockchain data.

Anyone that wants to build an application, one of these decentralized apps on top of blockchain data has a couple of options. They can either build their own indexing server and deploy it to somewhere like AWS. That takes away the whole idea of decentralization because then you have a single point of failure again. You can query data directly from the blockchain, from your client application, which takes a very long time. Both of those are not, I would say the most optimal way to build. But also if you're building your own indexing server, every time you want to come up with a new idea also, you have to think about the resources and time that go into it. Basically, I want to come up with a new idea and test it out, I have to basically build a server index, all this data, create APIs around it. It's time-intensive.

What the Graph protocol allows you to do is, as a developer you can basically define a subgraph using YAML, similar to something like cloud formation or a very condensed version of that maybe more Serverless Framework where you're defining, I want to query data from this data source, and I want to save these entities and you deploy that to the network. And that subgraph will basically then go and look into that blockchain. And will look for all the transactions that have happened, and it will go ahead and save those and make those available for public retrieval. And also, again, one of the things that you might think of is, all of this data is public. All of the data that's on the blockchain is public.

Jeremy: Right. Right. All right. Let me see if I could repeat what you said and you tell me if I'm right about this. Because this was one of those things where blockchain ... you're right. To me, it had a negative connotation. Why would you use the blockchain, unless you were building your own cryptocurrency? Right. That just seemed like that's what it was for. Then when AWS comes out with QLDB or they announced that or whatever it was. I'm like, okay, so this is interesting, but why would you use it, again, unless you're building your own cryptocurrency or something because that's the only thing I could think of you would use the blockchain for.

But as you said, with these blockchains now, you have highly sensitive transactions that can be public, but a real estate transaction, for example, is something really interesting, where like, we still live in a world where if Bank of America or one of these other giant banks, JPMorgan Chase or something like that gets hacked, they could wipe out financial data. Right? And I know that's backed up in multiple regions and so forth, but this is the thing where if you're doing some transaction, that you want to make sure that transaction lives forever and isn't manipulated, then the blockchain is a good place to do that. But like you said, it's expensive to write there. But it's even harder to read off the blockchain because it's that ledger, right? It's just information coming in and coming in.

So event storming or if you were doing event sourcing or something that, it's that idea. The idea with these indexers are these basically separate apps that run, and again, I'm assuming that these protocols, their software, and things that you don't have to build this yourself, essentially you can just deploy these things. Right? But this will read off of the blockchain and do that aggregation for you and then make that. Basically, it caches the blockchain. Right? And makes that available to you. And that you could deploy that to multiple indexers if you wanted to. Right? And then you would have access to that data across multiple providers.

Nader: Right. No single point of failure. That's exactly right. You basically deploy a very concise configuration file that defines how you want your data stored and made available. And then it goes, and it just starts at the very beginning and it queries all those blocks or reads all those blocks, saves the data in a database, and then it keeps up with additional new updates. If someone writes a new transaction after that, it also saves that and makes it available for efficient retrieval. This is just for blockchain data. This is the data layer for, but it's not just a blockchain data in the future. You can also query from IPFS, which is a file storage layer, somewhat S3. You can query from other chains other than Ethereum, which is kind of like the main chamber.

In the future really what we're hoping to have is a complete API on top of all public data. Anybody that wants to have some data set available can basically deploy a subgraph and index it and then anyone can then essentially query for it. It's like when you think of public data, we're not really used to thinking of data in this way. And also I think a good thing to talk about in a moment is the types of apps that you can build because you wouldn't want to store private messages on a blockchain or something like that. Right? The types of apps that people are building right now at least are not 100% in line with everything. You can't do everything I would say right now in Web3 that you can do in Web2.

There are only certain types of applications, but those applications that are successful seem to be wildly successful and have a lot of people interested in them and using them. That's the general idea, is like you have this way to basically deploy APIs and the technology that we use to query is GraphQL. That was one of the reasons that I became interested as well. Right now the main data sources are blockchains like Ethereum, but in the future, we would like to make that available to other data sources as well.

Jeremy: Right. You mentioned earlier too because there are apps obviously being built on this that you said are successful. And the problem though, I think right now, because I remember I speculated a little bit with Bitcoin and I bought a whole bunch of Ripple, so I'm still hanging on to it. Ripple XPR whatever, let's go. Anyways, but it was expensive to make a transaction. Right? Reading off of the blockchain itself, I think just connecting generally doesn't cost money, but if you're, and I know there's some costs with indexers and that's how that works. But in terms of the real cost, it's writing to the blockchain. I remember moving some Bitcoin at one point, I think cost me $30 to make one transaction, to move something like that.

I can see if you're writing a $300,000 real estate transaction, or maybe some really large wire transfer or something that you want to record, something that makes sense where you could charge a fee of $30 or $40 in order to do that. I can't see you doing that for ... certainly not for web streaming or click tracking or something like that. That wouldn't make sense. But even for smaller things there might be writing more to it, $30 or whatever that would be ... seems quite expensive. What's the hope around that?

Nader: That was one of the biggest challenges and that was one of the reasons that when I first, I would say maybe even considered this as a technology back in the day, that I would be considering as something that would possibly be usable for the types of applications I'm used to seeing. It just was like a no-brainer, like, no. I think right now, and that's one of the things that attracted me right now to some of the things that are happening, is a lot of those solutions are finally coming to fruition for fixing those sorts of things. There's two things that are happening right now that solve that problem. One of them is, they are merging in a couple of updates to the base layer, layer one, which would be considered something like Ethereum or Bitcoin. But Ethereum is the main one that a lot of the financial stuff that I see is happening.

Basically, there are two different updates that are happening, I think the main one that will make this fee transactional price go down a little bit is sharding. Sharding is basically going to increase the number of, I believe nodes that are basically able to process the transactions by some number. Basically, that will reduce the cost somewhat, but I don't think it's ever going to get it down to a usable level. Instead what the solutions seem to be right now and one of the solutions that seems to actually be working, people are using it in production really recently, this really just started happening in the last couple of months, is these layer 2 solutions. There are a couple of different layer 2 solutions that are basically layers that run on top of the layer one, which would be something like Ethereum.

And they treat Ethereum as the settlement layer. It's almost like when you interact with the bank and you're running your debit card. You're probably not talking to the bank directly and they are doing that. Instead, you have something like Visa who has this layer 2 on top of the banks that are managing thousands of transactions per second. And then they take all of those transactions and they settle those in an underlying layer. There's a couple different layer 2s that seem to be really working well right now in the Ethereum ecosystem. One of those is Arbitrum and then the other is I think Matic, but I think they have a different name now. Both of those seem to be working and they bring the cost of a transaction down to a fraction of a penny.

You have, instead of paying $20 or $30 for a transaction, you're now paying almost nothing. But now that's still not cheap enough to probably treat a blockchain as a traditional database, a high throughput database, but it does open the door for a lot of other types of applications. The applications that you see building on layer one where the transactions really are $5 to $20 or $30 or typically higher value transactions. Things like governance, things like financial transactions, you've heard of NFTs. And that might make sense because if someone's going to spend a thousand bucks or 500 bucks, whatever ...

Jeremy: NFTs don't make sense to me.

Nader: They're not my thing either, the way they're being, I would say, talked about today especially, but I think in the future, the idea behind NFTs is interesting, but yeah, I'm in the same boat as you. But still to those people, if you're paying a thousand dollars for something then that 5 or 10 or 20 bucks might make sense, but it's not going to make sense if I just want to go to an e-commerce store and pay $5 for something. Right? I think that these layer 2s are starting to unlock those potential opportunities where people can start building these true financial applications that allow these transactions to happen at the same cost or actually a lot cheaper maybe than what you're paying for a credit card transaction, or even what those vendors, right? If you're running a store, you're paying percentages to those companies.

The idea around decentralization comes back to this discussion of getting rid of the middleman, and a lot of times that means getting rid of the inefficiencies. If you can offload this business logic to some type of computer, then you've basically abstracted away a lot of inefficiencies. How many billions of dollars are spent every year by banks flying their people around the world and private jets and these skyscrapers and stuff. Now, where does that money come from? It comes from the consumer and them basically taking fees. They're taking money here and there. Right? That's the idea behind technology in general. They're like whenever something new and groundbreaking comes in, it's often unforeseen, but then you look back five years later and you're like, this is a no-brainer. Right?

For instance Blockbuster and Netflix, there's a million of them. I don't have to go into that. I feel this is what that is for maybe the financial institutions and how we think about finance, especially in a global world. I think this was maybe even accelerated by COVID and stuff. If you want to build an application today, imagine limiting yourself to developers in your city. Unless you're maybe in San Francisco or New York, where that might still work. If I'm here in Mississippi and I want to build an application, I'm not going to just look for developers in a 30-mile radius. That is just insane. And I don't use that word mildly, it's just wild to think about that. You wouldn't do that.

Instead, you want to look in your nation, but really you might want to look around the world because you now have things like Slack and Discord and all these asynchronous ways of doing work. And you might be able to find the best developer in the world for 25% or 50% of what you would typically find locally and an easy way to pay them might just be to just send them some crypto. Right? You don't have to go find out all their banking information and do all the wiring and all this other stuff. You just open your wallet, you send them the money and that's it. It's a done deal. But that's just one thing to think about. To me when I think about building apps in Web2 versus Web3, I don't think you're going to see the Facebook or Instagram use case anytime in the next year or two. I think the killer app for right now, it's going to be financial and e-commerce stuff.

But I do think in maybe five years you will see someone crack that application for, something like a social media app where we're basically building something that we use today, but maybe in a better way. And that will be done using some off-chain storage solution. You're not going to be writing all these transactions again to a blockchain. You're going to have maybe a protocol like Graph that allows you to have a distributed database that is managed by one of these networks that you can write to. I think the ideas that we're talking about now are the things that really excite me anyway.

Jeremy: Let's go back to GraphQL for a second, though. If you were going to build an app on top of this, and again, that's super exciting getting those transaction fees down, because I do feel every time you try to move money between banks or it's the $3 fee, if you go to a foreign ATM and you take money out of an ATM, they charge you. Everybody wants to take a cut somewhere along, and there's probably reasons for it, but also corporate jets cost money. So that makes sense as well. But in terms of the GraphQL protocol here, so if I wanted to build an application on top of it, and maybe my application doesn't write to the blockchain, it just reads from it, with one of these indexers, because maybe I'm summing up some financial transactions or something, or I've got an app we can look things up or whatever, I'm building something.

I'm querying using the GraphQL, this makes sense. I have to use one of these indexers that's aggregating that data for me. But what if I did want to write to the blockchain, can I use GraphQL to do a mutation and actually write something to the blockchain? Or do I have to write to it directly?

Nader: Yeah, that's actually a really, really good question. And that's one of the things that we are currently working on with the Graph. Right now if you want to write a transaction, you typically are going to be using one of these JSON RPC wallets and using some type of client library that interacts with the wallet and signs the transaction with the private key. And then that sends the transaction to the blockchain directly. And you're talking to the blockchain and you're just using something like the Graph to query. But I think what would be ideal and what we think would be ideal, is if someone could use a single technology, a single language, and a single abstraction to do everything, not only with reading and writing but also with subscriptions for real-time updates.

That's where we think the whole idea for this will ultimately be, and that's what we're working on now. Right now you can only query. And if you want to write a transaction, you basically are still going to be using something like ethers.js or Web3 or one of these other libraries that allows you to sign a transaction using your wallet. But in the future and in fact, we're already building this right now as having an end-to-end GraphQL library that allows you to write transactions as well as read. That way someone just learns a single API and it's a lot easier. It would also make it easier for developers that are coming from a traditional web background to come in because there's a little bit of learning curve for understanding how to create one of these signed providers and write the transaction. It's not that much code, but it is a new way of thinking about things.

Jeremy: Well I think both of us coming from the serverless space, we know that new way of thinking about things certainly can throw a wrench in the system when a new developer is trying to pick that stuff up.

Nader: Yeah.

Jeremy: All right. So that's the blockchain side of things with the data piece of it. I think people could wrap their head around that. I think it makes a lot of sense. But I'm still, the decentralized, the other things that you talked about. You mentioned an S3, something that's sort of an S3 type protocol that you can use. And what are some of the other ones? I think I've written some of them down here. Acash was one, Filecoin, Livepeer. These are all different protocols or services that are hosted by the indexers, or is this a different thing than the indexers? How does that work? And then how would you use that to save data, maybe save some blob, a blob storage or something like that?

Nader: Let's talk about the tokenomics idea around how crypto fits into this and how it actually powers a protocol like this. And then we'll talk about some of those other protocols. How do people actually build all this stuff and do it for, are they getting paid for it? Is it free? How does that work and how does this network actually stay up? Because everything costs money, developers' time costs money, and so on and so forth. For something like the Graph, basically during the building phase of this protocol, basically, there was white papers and there was blog posts, and there was people in Discords talking about the ideas that were here. They basically had this idea to build this protocol. And this is a very typical life cycle, I would say.

You have someone that comes up with an idea, they document some of it, they start building it. And the people that start building it are going to be basically part of essentially the founding team you could think of, in the sense of they're going to be having equity. Because at the end of the day, to actually launch one of these decentralized protocols, the way that crypto comes into it, there's typically some type of a token offering. The tokens need to be for a network like this, some type of utility token to keep the network running in the future. You're not just going to create some crypto and that's it like, right? I think that's the whole idea that I thought was going on when in reality, these tokens are typically used for powering the protocol.

But let's say early on you have let's say 20 developers and they all build 5% of the system, whatever percentage that you want to talk about, whatever. Let's say you have these people helping out and then you actually build the thing and you want to go ahead and launch it and you have something that's working. A lot of times what people will do is they'll basically have a token offering, where they'll basically say, okay, let's go ahead and we're going to mint X number of tokens, and we're going to put these on the market and we're going to also pay these people that helped build this system, X number of tokens, and that's going to be their payment. And then they can go and sell those or keep those or trade those or whatever they would like to do.

And then you have the tokens that are then put on the public market essentially. Once you've launched the protocol, you have to have tokens to basically continue to power the protocol and fund it. There are different people that interact with the protocol in different ways. You have the indexers themselves, which are basically software engineers that are deploying whatever infrastructure to something like AWS or GCP. These people are still using these cloud providers or they're maybe doing it at their house, whatever. All you basically need is a server and you want to basically run this indexer node, which is software that is open source, and you run this node. Basically, you can go ahead and say, okay, I want to start being an indexer and I want to be one of the different nodes on the network.

To do that you basically buy some GRT, Graph Token, and in our case you stake it, meaning you are putting this money up to basically affirm that you are an indexer on the protocol and you are going to be accepting subgraph developers to deploy their subgraphs to your indexer. You stake that money and then when people use the API, they're basically paying money just like they might pay money to somewhere like API gateway or AppSync. Instead, they're paying money for their subgraph and that money is paid in GRT and it's distributed to the people in the ecosystem. Like me as a developer, I'm deploying the subgraph, and then if I have a million people using it, then I make some money. That's one way to use tokens in the system.

Another way is basically to, as an outside person looking in, I can say, this indexer is really, really good. They know what they're doing. They're a very strong engineer. I'm going to basically put some money into their indexer and I'm basically backing them as an indexer. And then I will also share the money that comes in from the query fees. And then there are also people that are subgraph developers, which is the stuff that I've been working with mainly, where I can basically come up with a new API. I can be like, it'd be cool if I took data from this blockchain and this file system and merged it together, and I made this really cool API that people can use to build their apps with. I can deploy that. And basically, people can signal to this subgraph using tokens. And when people do that, they can say that they believe that this is a good subgraph to use.

And then when people use that, I can also make money in that way. Basically, people are using tokens to be part of the system itself, but also to use that. If I'm a front end application like Uniswap and I want to basically use the Graph, I can basically say, okay, I'm going to put a thousand dollars in GRT tokens and I'm going to be using this API endpoint, which is a subgraph. And then all of the money that I have put up as someone that's using this, is going to be taken as the people start using it. Let's say I have a million queries and each query is one, 1000th of a cent, then after those million queries are up, I've spent $100 or something like that. Kind of similar to how you might pay AWS, you're now paying, you know, subgraph developers and indexers.

Jeremy: Right. Okay. That makes sense. So then that's the payment method of that. So then these other protocols that get built on top of it, the Acash and Filecoin and Livepeer. So those ...

Nader: They're all operating in a very similar fashion.

Jeremy: Okay. All right. And so it's ...

Nader: They have some type of node software that's run and people can basically run this node on some server somewhere and make it available as part of the network. And then they can use the tokens to participate. There's Filecoin for file storage. There's also IPFS, which is actually more of, it's a completely free service, but it's also not something that's as reliable as something like S3 or Filecoin. And then you have, like you mentioned, I believe Acash, which is a way to execute arbitrary code, business logic, and stuff like that. You have Ceramic Network, which is something that you can use for authentication. You have Livepeer which is something you use for live streaming. So you have all these ideas, these decentralized services fitting in these different niches.

Jeremy: Right, right. Okay. So then now you've got a bunch of people. Now you mentioned this idea of, you could say, this is a good indexer. What about bad indexers? Right?

Nader: That's a really good question.

Jeremy: Yeah. You're relying on people to take data off of a public blockchain, and then you're relying on them to process it correctly and give you back good data. I'm assuming they could manipulate that data if they wanted to. I don't know why, but let's say they did. Is there a way to guarantee that you're getting the correct data?

Nader: Yeah. That's a whole part of how the system works. There's this whole idea and this whole, really, really deep rabbit hole of crypto-economics and how these protocols are structured to incentivize and also disincentivize. In our protocol, basically, you have this idea of slashing and this is also a fairly known and used thing in the ecosystem and in the space. It's this idea of slashing. Basically, you incentivize people to go out and find people that are serving incorrect data. And if that person finds someone that's serving incorrect data, then the person that's serving the incorrect data is, quote-unquote, slashed. And that basically means that they're not only not going to receive the money from the queries that they were serving, but they also might lose the money that they put up to be a part of the network.

I mentioned you have to actually put up money to deploy an indexer to the network, that money could also be at risk. You're very, very, very much so financially disincentivized to do that. And there's actually, again, incentives in the network for people to go and find those people. It's all-around incentives, game theory, and things like that.

Jeremy: Which makes a ton of sense. That's good to know. You mentioned, you threw out the number, five years from now, somebody might build the killer app or whatever, they'll figure out some of these things. Where are we with this though? Because this sounds really early, right? There's still things that need to be figured out. Again, it's public data on the blockchain. How do you see this evolving? When do you think Web3 will be more accessible to the masses?

Nader: Today people are actually building really, really interesting applications that are fitting the current technology stack, what are the things that you can build? People are already building those. But when you think about the current state of the web, where you have something like Twitter, or Facebook or Instagram, where I would say, especially maybe something like Facebook, that's extremely, extremely complex with a lot of UI interaction, a lot of private data, messages and stuff. I think to build something like that, yeah, it's going to be a couple of years. And then you might not even see certain types of applications being built. I don't think there is going to be this thing where there is no longer these types of applications. There are only these new types. I think it's more of a new type of application that people are going to be building, and it's not going to be a winner takes all just like in all tech in my opinion.

I wouldn't say all but in many areas of tech where you're thinking of something as a zero-sum game where I don't think this is. But I do think that the most interesting stuff is around how Web3 essentially enables native payments and how people are going to use these native payments in interesting ways that maybe we haven't thought of yet. One of the ways that you're starting to see people doing, and a lot of venture capitalists are now investing in a lot of these companies, if you look at a lot of the companies coming out of YC and a lot of the new companies that these traditional venture capitalists are investing in, are a lot of TOMS crypto companies.

When you think about the financial incentives, the things that we talked about early on, let's say you want to have the next version of YouTube and you don't want to have ads. How would that even work? Right? You still need to enable payments. But there's a couple of things that could happen there. Well, first of all, if you're building an application in the way that I've talked about, where you basically have these native payments or these native tokens that can be part of the whole process now, instead of waiting 10 years to do an IPO for an application that has been around for those 10 years and then paying back all his investors and all of those people that had been basically pulling money out their pockets to take part in.

What if someone that has a really interesting idea and maybe they have a really good track record, they come out with a new application and they're basically saying, okay, if you want to own a piece of this, we're going to basically create a token and you can have ownership in it. You might see people doing these ICO's, initial coin offerings, or whatever, where basically they're offering portions of the company to anyone that wants to own it and then incentivizing people to basically use those, to govern how the application is built in the future. Let's say I own 1% of this company and a proposal is put up to do something new. I can basically say, I can use that portion of my ownership to vote on things. And then people that are speculating can say, this company is doing interesting things. I'm going to buy into it, therefore driving the price up or down.

Kind of like the same way that you see the traditional stock market there, but without all of the regulation and friction that comes with that. I think that's interesting and you're already seeing companies doing that. You're not seeing the majority of companies doing that or anything like that, but you are starting to see those types of things happening. And that brings around the discussion of regulations. Is ... can you even do something like that in the United States? Well, maybe, maybe not. Does that mean people are going to start building these companies elsewhere? That's an interesting discussion as well. Right now if you want to build an application this way, you need to have some type of utility that these tokens are there for. You can't just do them purely on speculation, at least right now. But I think it's going to be interesting for sure, to watch.

Jeremy: Right. And I think too that, I'm just thinking if you're a bank, right? And you maybe have a bunch of private transactions that you want to keep private. Because again, I don't even know how, I don't know how we get to private transactions on the blockchain. I could see you wanting to have some transactions that were public blockchain and some that were private and maybe a hybrid approach would make sense for some companies.

Nader: I think the idea that we haven't really talked about at all is identity and how identity works compared to how we're used to identity. The way that we're used to identity working is, we basically go to a new website and we're like, this looks awesome. Let me try it out. And they're like, oh wait, we need your name, your email address, your phone number, and possibly your credit card and all this other stuff. We do that over and over and over, and over time we've now given our personal information to 500 people. And then you start getting these emails, your data has been breached, every week you get one of these emails, if you're someone like me, I don't know. Maybe I'm just signing up for too much stuff. Maybe not every week, but maybe every month or two. But you're giving out your personal data.

But we're used to identity as being tied to our own physical name and address and things like that. But what if identity was something that was more abstract? And I think that that's the way that you typically see identity managed in Web3. When you're dealing with authentication mechanisms, one of the most interesting things that I think that is part of this whole discussion is this idea of a single sign-on mechanism, that you own your identity and you can transfer it across all the applications and no one else is in control of it. When you use something like an Ethereum wallet, like MetaMask, for example, it's an extension you can just download and put crypto in and basically make payments on the web with. When you create a wallet, you're given a wallet address. And the wallet address is basically created using public key cryptography, where basically you start with this private key, your public key is derived from the private key, and then your address is dropped from the public key.

And when you send a transaction, you basically sign the transaction with your private key and you send your public key along with the transaction, and the person that receives that can decode the transaction with the public key to verify that that's who signed the transaction. Using this public key cryptography that only you can basically sign with your own address and your own password, it's all stored on the blockchain or in some decentralized manner. Actually in this case stored on the blockchain or it depends on how you use it really, I guess. But anyway, the whole idea here is that you completely own your identity. If you never decide to associate that identity with your name and your phone number, then who knows who's sending these transactions and who knows what's going on, because why would you need to associate your own name and phone number with all of these types of things, in these situations where you're making payments and stuff like that. Right?

What is the idea of a user profile anyway, and why do you actually need it? Well, you might need it on certain applications. You might need it or want it on social network, or maybe not, or you might come up with a pseudonym, because maybe you don't want to associate yourself with whatever. You might want to in other cases, but that's completely up to you and you can have multiple wallet addresses. You might have a public wallet address that you associate your name with that you are using on social media. You might have a private wallet address that you're never associating with your name, that you're using for financial transactions. It's completely up to you, but no one can change that information. One of the applications that I recently built was called Decentralized Identity. I built it and release it a few days ago.

And it's an implementation of this and it's using some of these Web3 technologies. One of them is IDX. One of them is Ceramic, which is a decentralized protocol similar to the Graph but for identity. And then it's using something called DIDs, which are decentralized identifiers, which are a way to have a completely unique ID based off of your address. And then you own the control over that. You can basically go in and make updates to that profile. And then any application across the web that you choose to use can then access that information. You're only dealing with it stored in one place. You have full control over it, at any time you can go in and delete that. You can go in and change it. No one has control over it except for you.

The idea of identity is a mind-bending thing in this space because I think we're so used to just handing everybody our real names and our real phone numbers and all of our personal information and just having our fingers crossed, that we're just not used to anything else.

Jeremy: It's all super interesting. You mentioned earlier about, would it be legal in the United States? I'm thinking of all these recent ransomware attacks and I think they were able to trace back some Bitcoin transaction, they were actually able to trace it back to the individual group that accepted the payment. It opens up a whole can of worms. I love this idea of being anonymous and not being tracked, but then it's also like, what could bad actors do with anonymous financial transactions and things like that? So ...

Nader: There kind of has been anonymous transactional layer for a long time. Cash brought in, you can't really do a lot of illegal stuff these days without cash. So should we get rid of cash? I think with any technology ...

Jeremy: No, but I mean, there's a limit though, right? You can't withdraw more than $10,000 worth of cash without the FBI being flagged and you can't deposit more, you know what I mean?

Nader: You can't take a million dollars worth of Bitcoin that you've gotten from ransomware and turn it into cash either.

Jeremy: That's also true. Right.

Nader: Because it's all tracked on the blockchain, that's probably how they caught those people. Right? They somehow had their personal information tied to a transaction, because if you follow these transactions long enough, you're going to find some origination point. I agree though. There's definitely trade-offs with everything. I don't think I'm ever the type to argue that. There's good things and there's bad things. I think you have to look at the whole picture and decide for yourself, what you think. I'm the type that's like, let's lay out all of the ideas and let the market decide.

Jeremy: Right. Yeah. I totally agree with that. All this stuff is fascinating, there is way too much more for me to learn at this point. I think my brain is filled at this point. Anything else about Edge & Node? Any cool things you're working on there or anything you want people to know?

Nader: We're working on a couple of different projects. I can't really talk about some of them because they're not released yet, but we are working on a new version of something called Everest, and Everest is already out. If you want to check it out, it's at everest.link. It's basically a repository of a bunch of different applications that have already been built in the Web3 ecosystem. It also ties in a lot of the stuff that we talked about, like identity and stuff like that. You can basically sign in with your Ethereum wallet. You can basically interact with different applications and stuff, but you can also just see the types of stuff people are building. It's categorized into games, financial apps. If you've listened to this and you're like, this sounds cool, but are people actually building stuff? This is a place to see hundreds of apps that people have are already built and that are out there and successful.

Jeremy: Awesome. All right. Well, listen, Nader, this was awesome. Thank you so much for sharing this with me. I know I learned a ton. I hope the listeners learned a ton. If people want to learn more about this or just follow you and keep up with what you're doing, what's the best way to do that?

Nader: I would say check out Twitter, we're on Twitter @dabit3 for me, @edgeandnode for Edge & Node, and of course @graphprotocol for Graph protocol.

Jeremy: Okay. And then edgeandnode.com. Your YouTube channel is just youtube.com/naderdabit, N-A-D-E-R D-A-B-I-T. And then you had an article on Web3 and I'll put it in the show notes.

Nader: Yeah. Put it in the show notes. For freeCodeCamp, it's called what is Web3. And it's really a condensed version of a lot of the stuff we talked about. Maybe go into a little bit more depth around native payments and how people might build companies in the way that we've talked about here.

Jeremy: Awesome. All right. Well, I will get all that stuff into the show notes. Thanks again, Nader.

Nader: Thanks for having me. It was good to talk.