Lessons from 10 Years of the Ethereum Identity Stack

Transcript

[applause] >> Hello everybody. Very glad to be here. Uh my name is Brantley Milligan and I'm going to be talking about lessons from 10 years of the Ethereum identity stack. Deep into identity. So who am I? This is my on-chain uh profile. My on-chain name ENS name is brantley.eth. How many people here have a .eth name? Raise your hand. You have a couple people here. Okay, if you don't have a .eth name, you got to go to ens.domains and you got to get one. I'm going to talk about why that's important.

Um It's my on-chain uh Ethereum profile. I'll talk about what that is. So the story of the Ethereum identity stack actually goes back about 50 years to the 1970s when um we had one of the most important uh discoveries, you could say, or inventions uh in computer science of our lifetime or, you know, of of this generation, which is public key cryptography. How many people here know what public key cryptography is? Okay, raise your hand. It is extremely powerful.

Um [snorts] it allows a person to do encryption with another person without having to trade um a secret decryption key ahead of time. You can actually just trade a public key, you can encrypt a message for that person, and only that message can decrypt it without you having to uh traded a decryption key ahead of time. This had never been possible before in all of human history. It was invented in the 1970s. This was considered to be so powerful that the US government classified it and said it's illegal to use in most public products.

And so it was invented in the 1970s, it was used only by the US government for a period of time. Um but then in the 1990s, there was a guy named Phil Zimmerman and a group of people called the Cypherpunks who they lobbied to make to make it legal for the public to use and they were successful. This was extremely important for the modern world. We would not have e-commerce, we would not have so many things on the internet unless we had public key cryptography. So like uh in your browser when it says HTTPS, that S there is for SSL, that's a type of cryptography that is um allowed because of the work they did in the 1990s.

So the Cypherpunks realized that public key cryptography could be the basis um of many things including digital identity. Uh but they had a hard time getting adoption because uh you know, a lot of people believe in privacy, they believe in liberty, but most people even if they believe in that are not motivated enough to change their life accordingly. They need like financial incentive or fun or something else. So um there was a a version of public key cryptography that was meant for people called PGP. But you know, it was adopted by like you know, techies or something but never really kind of broke out of that.

Flash forward to 2008. Um we had the creation or the proposal of the creation of Bitcoin. And it used public key cryptography for wallets. Ahem. >> [clears throat] >> This was very important cuz it was like another use case for public key cryptography. 2011, we had Namecoin which was the first altcoin. And essentially the uh the insight from Namecoin was that um blockchain technology was invented for Bitcoin for money, but we could use this for other things including for naming and identity. So that Namecoin Namechain, I'm sorry, not Namechain, Namecoin was its own blockchain chain dedicated to um names uh specifically dot bit names.

And uh it was innovative, it was a cool thing, never really got adoption, but the uh the seeds of the idea was there. In 2014, uh Vitalik published the Ethereum white paper. He said there were five areas um uh for use cases of Ethereum, and one of them was identity, right there. He actually says um we could use Ethereum to have a better blockchain-based naming system, maybe even better than Namecoin. 2015, of course, Ethereum launched. It also used public key cryptography for wallets. Um and Ethereum, of course, can run any smart contract. So, here's the key thing. What what Bitcoin and then Ethereum in particular did is it got tons of people private keys who

otherwise would never have had it. Right? So, crypto has accomplished with economic incentives what Cypherpunks couldn't, with appeals to privacy and freedom. The widespread adoption of private keys and the flourishing of a private key management industry, which we call wallets. Uh wallets are just private key management systems. Those have existed for years, but they always sucked. They were really bad. Uh crypto wallets, you think maybe they suck today or could be better? They're way better than anything we had until just a few years ago. Yes, they need to keep getting better, but this is happening because we have an economic incentive. It's just not It's not simply an appeal to, you know, values or something, which motivates some people, but not enough people.

So, Vitalik said, "Okay, with Ethereum, we could do identity." There were a lot of early attempts at doing um identity on Ethereum early on, and a lot of them failed. I think the reason a lot of them failed is because they thought of identity or even socialize a single app. And uh turns out identity is not a single problem. It's a a single hard problem. It's like 20 hard problems. And if you try to solve all of them at once, you might solve five of them if you're really smart team, but you're probably not going to be able to solve all of them. You're going to cut corners. You're not going to get really what we want, and your project's going to fail.

And uh this is a key thing that I've learned from from 10 years uh in the Ethereum identity stack, which is this. Building things in a decentralized is way harder than building things in a centralized way. And yes, there are many Web 3 devs who wish they were building Web 2 because their life would be a lot easier. Which is fine. I've nothing against Web 2. If you want to do Web 2, go do it. But if you're going to be in crypto, the whole point is we're trying to do things in a decentralized way. So then, the next big breakthrough was Ethereum Name Service, ENS. Development started in 2016, launched in 2017. Uh and it began as dot eth names for decentralized websites. It was modeled after DNS. Very soon after, they said, "Well, we can also use this for crypto

payments." Interestingly, when it launched, nobody thought of it as identity primitive. People were not using that terminology. It was actually thinking of it more from this DNS analogy. But then in 2021, uh 4 years after launch, I was on the team at that point, and I realized that ENS was actually accidentally your crypto username and profile. People were starting to store profile data. People were starting to put it on their Twitter to represent themselves. And um I remember writing a thread about this in early 2021, and I for for the first time wrote out, "It's like your you know, Web 3 username." And like hit me like, "Oh my gosh, yeah. That is That's what we're really doing here."

And I think username is actually the meta use case for all the other things. So, uh you want to send me money, you know, for crypto payments, well, send it to me. Send it to my username. You want to go to my decentralized website, you just put my name in. You know, you can see like my page on the decentralized web. I got to say, ENS um has been very successful so far. And uh let's not take for take this for granted. It's really amazing how far it's come so far. There have been many, many, many attempts at blockchain-based naming. Obviously, I mentioned uh Namecoin, but there have been dozens and dozens of other attempts, none of most of which you haven't heard of because they tried and they failed. It's really hard to have success in these types of things.

And um so, why has ENS succeeded when others didn't? A couple of reasons. One, it's not trying to compete directly with DNS. ENS sees itself as a complementary protocol that can expand um, the internet name space. Other blockchain-based naming systems have tried to directly compete with DNS, which I think is a terrible mistake because you're not going to beat DNS because it's extremely entrenched. Um, another thing is ENS has great engineering. It's had great engineers right from the very beginning. Don't take that uh, for granted. Number three, long uh, long time horizons in protocol design. So, ENS is a It's not just a science project. It's a real attempt at a new protocol of the internet, which means from the beginning, the creators, the developers

of it, not, you know, from the beginning and then ongoing, have made decisions for its long-term good rather than for short-term, you know, just hype or, you know, extraction. Um, you know, yes, you can have some short-term success with that, but then it flares away and then you don't have anything. ENS has always been for the long-term and it's already been around for over 9 years, which makes it extremely Lindy. It's one of the longest-running apps without any kind of um, uh, without any uh, interruption on Ethereum. And uh, you know, it's here to stay in the long-term.

So then also in 2021, I realized that the Ethereum community had accidentally solved of uh, or created a viable solution to the single sign-on problem. So, the single sign-on problem is uh, we have, you know, tons of apps that you want to use, but every single one has its own sign-in system, which is really annoying, right? So, one solution to this has been, well, you can sign in with Google everywhere, which is okay, but then it's like, well, now Google controls your life. You don't like that. So, it'd be great, wouldn't it be great if the internet had a decentralized, user-controlled single sign-on system where you had your account for the internet, you had a username and a profile for the internet, and then you brought that with you to

all the different apps. So, you're you don't have to get a new username, you don't have to create a new profile every time, you don't have to make a new password. You bring your own account to the app. Well, that is happening with sign-in with Ethereum. Sign with Ethereum, it's EIP-4361. It's an authentication standard for Ethereum accounts, pairs well with ENS, so your Ethereum account can be your um your account for the internet. And ENS can be your username and profile. This is supported in uh wallets like MetaMask and many others and throughout the ecosystem.

Um another uh project that I've been involved in is called Ethereum Photo Protocol. So, you know, sign with Ethereum allows you to log in anywhere with your Ethereum account, ENS is your uh crypto username and profile. Ethereum Photo Proto- Protocol is your on-chain social graph protocol for Ethereum accounts, um that you can follow other Ethereum accounts on-chain, and again, take that with you between apps. It's on-chain, it's open for anybody to use, has tens of thousands of users, and it complements ENS and sign with Ethereum very well. In talking about the Ethereum identity stack, there's lots of other things I could talk about. Some of these you may have heard of. Ethereum Attestation Service, XMTP, Ethereum Common Protocol, many more. I'm not attempting to be um

exhaustive in this presentation, um but, you know, there are many parts here that you could talk about. Another project that uh my team has made is Ethereum Identity Kits, which is just a component a React component library that makes it very easy to integrate ENS, sign with Ethereum, and other things into your app. It's at ethidentitykit.com. So, what is Ethereum identity? This is what I think after having uh worked on this for about 10 years. Ethereum identity is not a single app. Instead, it is the sum of the information and relationships held in various decentralized protocols and open standards, which I call the Ethereum identity stack.

The Ethereum identity stack is itself decentralized, rather than being vertically integrated. I think this is a good thing. In other words, we've unbundled the elements of identity to their smallest viable parts and have separated and have separate teams solving them individually with decentralized protocols and open standards. What this means is that each part can stand or fall on its own merits, rather than being an all-or-nothing proposition. So, for example, if you like ENS and you're all in on that, but you don't like the sign with Ethereum standard EIP-4361, you can make your own authentication standard that does not impact ENS or vice versa.

So, it means that at any point, people can improve a part of the stack or replace a part of the stack without giving up the entire stack. That makes the Ethereum identity stack much more resilient and I think much better in the long term. Here's something I want to emphasize to end. To get to where we are today with ENS, the sign with Ethereum, with these other elements okay? This has been a long road, very difficult to do. It's easy to take these things for granted. Like I said with ENS, there have been many attempts at naming systems and usernames and all these things like this, the vast majority of which have failed. ENS by with luck and the right people and the right circumstances at the right times

has managed to get to where it is and it has had, you know, you know, kind of pretty good success and it needs to keep growing. Um but to even get to that point um is really kind of like miraculous and uh we should not take that for granted. Same thing with sign with Ethereum, really hard to coordinate a lot of different people, you know, wallets and apps around the industry to use the same standard. We're not 100% there, but we've made a lot of progress and there's lots uh work still to be done. But, because of all this hard work over the last 10 years, we do finally have these key parts of the stack.

That is the end of my presentation. If you have questions, I am free to discuss after this presentation. Thank you very much. >> [applause]