Sylve on borders, truth, and hardness with House of ZK

Sylve was again invited to the House of ZK podcast to chat with Alice Liu, education lead at ZKM, and Ming Guo, chief scientist at ZKM. They discussed trustlessness, borders on the Internet, and what ZK brings to our shared cultures and relationships.

Watch the video or read our highlights below!

Blockchain and allowing people to create truth

When you start looking at something like Bitcoin, you can say, "These transactions are immutable." A transaction that goes through is done, and there's no way to go back in time. And then you start questioning: it doesn't seem that there are a lot of things in the world that have this property of what Josh Stark calls **hardness,** things that don't change. It has this property of immutability that is just as trustworthy and eternal as gravity.

I think if I hadn't gone into technology, I’d have practiced law. The fun thing with the law is that it's absolutely everywhere. Every single interaction that you have is regulated, and there is a contract for it, and everything is mediated through that. The fun thing that I realized in my law classes is that the law is very monopolistic, and that has not always been the case.

In Europe, in particular, we love to think about the Middle Ages as a horrible time when everyone was under a sovereign king, but it was a very decentralized time.

One of my favorite pieces of trivia from the 13th or 14th century was that there were many competing standards for law and also competing standards for coinage. So, people were selling their money, and because they were trusted merchants, they would go around different fairs and villages and sell their coins. And because they were reputable merchants, you’d trust their coinage. You trust that they actually put the right amount of silver or gold inside their coins.

I've always been very interested in how you allow people to create norms and how you allow people to contractualize. The first time I truly realized what blockchain could give us was, oddly enough, NFTs.

The art market is regulated by intellectual property law. In order for you to sell something, it needs to be represented as something that you can commercialize within the realm of IP law or in the art market. And when I studied the art market, I was looking at an art piece, which was a video. And I thought, okay, that's actually pretty cool. It was a looped gif of a guy going up a pool and then jumping into the pool.

And I was wondering: “How do I buy that? How does that work?”

I realized that in order to buy it, you needed to turn it into an asset. And for that to become an asset, you need a whole installation: a physical TV that would hold it, for instance. You were not able to sell the file. You had no way of doing it.

NFTs tremendously lower the cost of things to be contractualized or commercialized.

And suddenly, you went from, “This thing has to cost like 4K because you need a whole installation and you need to talk to a lawyer,” to “It costs five bucks to go onchain and mint this”. You’ve tremendously lowered the cost of creating norms.

Blockchains are truth-making and norm-making machines, and that's a desirable outcome. Zero-knowledge extends that property even more.

The power of attested data

I recommend reading The Mystery of Capital by economist Hernando de Soto. It analyzes and asks why capitalism isn't always the solution. In some countries where you had sweeping liberalization actions in South America, it didn't lift poverty. But capitalism is a great tool: you can use capital, you can have a house, and you can use it as collateral to get credit. Then, you can buy new things and start a business. Why didn't that work?

He stated that some things were not considered within the realm of commerce. So he went around a favela in South America, and he asked, " Who does this house belong to?”

He realized that there are informal ways of knowing how much a house is worth and who owns it. This works because it's local. The 200 people living in that neighborhood all agree that that person owns that house. But if you go a mile away, people won’t know that. And if I look at the government records, nothing tells me whose house it is, and therefore, I cannot trust that.

The book said that if you had a way to represent these intersubjectivities, like these networks of truth, link them with each other, and make sure that they can be integrated into something bigger, you could have access to more capital. And this could be very useful. But this is currently dead capital, things that cannot be used outside of their own little realm. That's the great thing about the dollar, for example. It’s a fantastic API for value: people are always going to accept the dollar. It's much more difficult to use your favela house as collateral.

One way that we solve this problem of intersubjectivity is to put everything onchain. If everything runs on the blockchain, then there's no problem because everything is laid bare. And it's transparent, and therefore, you can create DeFi about it.

But the past 15 years of blockchains have shown us that we can’t put everything onchain. It's not performant enough. And some things were never bridged. There are still things that are going to stay in the real world.

What I really love about ZK is that it can use attested data to produce statements about them. And so if 200 people decide to trust that this house belonged to someone, they can collectively sign this fact. And if you trust that group, you're able to extend that fact somewhere else. Like, I have proof that this person has this house, attested by this group of people. And you're able to have things like transitive trust: if I trust you 80% and you trust that person 80%, then I will trust them by the factor of 80% of 80%.

And that's really cool!

That's why I talk about 50 shades of trust, having this ability to have things that are not truthful and not trustless. You've got something in between, and that's really neat.

Shifting to a different truth model

I remember an old The Economist article saying that the worst fate for a country is not to be like Greece and fail overnight, but rather to be like Argentina, one of the most powerful countries and best places to live on earth at the beginning of the 20th century and then brutally collapsing because of hyperinflation and cronyism.

I like blockchain because it's an engine for pluralism. This means that you can use different methods and speed up the rate of experimentation with currency.

Whenever I talked about cryptocurrencies to my friends about 10 years ago, they would always answer, « Why do I need this? I have PayPal. » And I replied, «You have Paypal because of 400 years of legal history; you have banks, and you have IT systems that make it possible. What blockchain allows you to do is rebuild this whole thing in a few thousand lines of code.»

And that is incredible! If you're able to speed-run 400 years of commerce to get to the same point, what else can you build on top of that?

That has led some people to think that we've solved the problem. They think we're just going to have this machine that is never going to change; there will be 21 million Bitcoins, and everything should run on Bitcoin.

That's not a good idea.

I think that's also very crude intellectually: things are more complex than “no inflation, no problem.”

Bitcoin was born out of the cypherpunk movement. There have always been a lot of libertarian thoughts in there, and I also felt that way when I was 15.

But then I grew up and realized that it's great that there's democratic oversight over a currency. It is great that, in some instances, you need to devalue. It is great that sometimes you can change the flow of money.

Obviously, it's a tool, and like any tool, it can be used in bad ways.

But the result was that in 2008, we were so scared of this lever being used the wrong way that we just decided to remove it altogether. And so we ended up with trustlessness. We thought we had solved all the issues, so we created this tiny little island in the middle of the Pacific where everything was well and dandy. And we thought that if everyone moved over to this tiny little island in the middle of the Pacific, everything would be great.

The past 15 years have shown us that no one wants to go live on a tiny little island in the Pacific.

People want to live where they belong − in Berlin, in Hong Kong, wherever they are − and use the currencies that they have. That's why there are so many debates about whether people should use ETH or native cryptocurrencies rather than stablecoins.

There was a standard called ERC20C, proposed by a team at Stanford about one or two years ago. The idea was to allow token transfers that could roll back, which is a great feature. I would give this to my parents! I would even have this for myself. I would be very happy to have a way to call someone and say, «Hi, sorry, I've made a mistake. I want to roll it back. »

They got lambasted. They were told that this is not what blockchain should be about. Blockchain should be about pure, absolute trustlessness. And if you make a mistake, that's your problem, which is a statement that people usually have until they make a mistake and send $2,000 to a random address, and they can never get back. It is great to have these fail-safes in place.

And that's what's great about cryptocurrency.

You have the option of trustlessness, but I strongly believe that the baseline shouldn't necessarily be that. It should be an option; that’s it. Not everyone is going to want to wear the tinfoil hat. If you're very happy with the hardware wallet and the 24 words buried in your backyard, good for you. But most people are going to want to use banks.

ZK can interact with the real world

The core problem of blockchain is that it's a great norm-making machine, but it's a very short-sighted machine that can only understand whatever the machine understands.

Whatever is on the tiny island in the Pacific can be reasoned about, can be computed over, and that's fantastic.

But that only works for blockchain-native things.

How do we make sense of this with the rest of the value and the rest of the information that is outside of the blockchain? It's a very isolated environment that understands itself. But how does this relate to the rest of the world?

So you end up with bridges and oracles, and it becomes a nightmare.

ZK really shines in making sure that it does not add more trust assumptions to the attestation. So, if I have a passport and I am able to produce zero-knowledge proof that I'm a French citizen, that's a cryptographic statement. The proof does not add more trust assumptions to that. And that's something I can directly put onchain. That's something that the blockchain machine is able to parse, understand, and use without adding any other trust assumption. That's a very good thing.

It's directly from attestation to the blockchain machine, and that's really, really useful. That's the most useful thing that we have for ZK.

Obviously, if you have a shit attestation, it is going to be a shit proof. The proof itself is not going to change anything. If the government's key leaks and you end up with a fake passport, you will be able to create fake proofs. That's a given. But that's a good thing. It means there are no additional trust assumptions: garbage in, garbage out. But if you trust the original attestation, then you're good to go.

Challenges of going onchain

One thing that I've realized with ZK is that it makes it very easy to create boundaries.

For example, at DevCon, you could get a discount if you could prove that you were from a Southeast Asian country. So you would generate proof from your passport that you're from a Southeast Asian country and you're able to get in. It's a way to delineate a circle and decide who's in and who's out. And that's for a really good reason. And honestly, that was a really perfect example: to do otherwise, you would need costly KYC.

Zero-knowledge reduces the cost of discrimination. And when you start looking at it this way, that doesn't sound so great. Does it mean I'm able to discriminate against people more easily? Is the marginal cost of racist acts going to zero?

ZK extends attestations and signatures. If we're going into a world where you start to sign more things, it means that you can extend them, reason over them, and commercialize them.

For example, in France, we have a signature standard called 2D doc, which signs many official documents. So your tax income reports and utility bills are signed, which is really good. You can attest to their veracity, which is really useful.

But then you can get into a situation where, in order to get to an online service in France, they're going to be geofenced. And you will constantly have to prove that you can be let into the inside group.

We have the technology to make that more palatable or easier. But if we start signing more things, we may end up in a situation where you're just going to have to constantly prove and show something that makes it so that you can be let in the circle.

This ability to delineate and create boundaries is fantastic. It's really great. It's a great tool. But at the same time, when you really lay it out, it's a way to discriminate more easily. That's pretty fucked.

Discoverability and borders on the Internet

Here’s a line of reasoning that I get from both Arnaud Schenk from the Aztec team and the Tonk.gg team, Baz and Goblin Oats.

I reread John Perry Barlow's Declaration of Independence of Cyberspace, which was written in the 1980s in Davos.

What he says is very telling: this guy in Davos declares that we're independent, like a completely different country, and that laws do not bind us. He also uses a metaphor where we have no bodies. We exist in this ether, but unintended of the internet, and there, we don't have a body. There's like one global consciousness.

There's a whole body of research about the fact that a lot of the early internet thoughts were really embedded in this new-age idea of being connected with this giant World Identity. It’s obvious when you read that text that they really believed it would erase all the differences between all the people.

They thought that if you could speak with someone on the Internet, connect, and share information, then all the problems would be solved.

Forty years of the Internet later, you realize that if you put people on the Internet, they argue and send slurs at each other, which is pretty bad.

It's weird to say that this was a completely unforeseen event. We never thought that helping people connect with no sense of physicality means that there are no customs.

Discoverability on the Internet

There isn't much discoverability on the Internet.

You're either completely anonymous, or you're posting your entire life on Instagram. It isn't easy to have layers of interaction with someone.

Something that I love about the Tonk team is that they're looking at ZK and cryptography as a way to mediate experiences between people on the internet the same way you meet people in real life.

If you go into a bar, you can't hear everyone's conversations.

You can see vaguely in the background that there's maybe two people on a date, then there's a group of four people over there, and then you start talking to maybe one person, and they reveal some things about themselves, and they're like, you know what? Come with my buddies; I'm going to buy you a beer. And then you get led into this other group.

If you take this metaphor for how the internet works, the first possibility is zero discoverability, which is group chats. For example, I don’t know where the ZKM Slack server is or how it can be let in. There's no discoverability for me about this.

On the other hand, I can go on Reddit or Twitter and then have direct access to 6 billion people simultaneously.

But there's not really this idea of « I could let these people into my group » or « I could find this group with these people » because sometimes you don't even know the group exists. There are layers of interaction between people, which comes back to the thing we discussed about boundaries. I think it's interesting.

I’ve been wondering whether there are alternatives. An example of this is alternative, decentralized designs for Twitter, like Farcaster and BlueSky. But these alternatives are predicated on the fact that everything is public at all times.

It's much more difficult to build a decentralized dating app.

The reason is that you have data that has to be private or data that has to be revealed to a specific group of people. Because I do not want the world to know that I'm a Linkin Park fan, but I would really love to meet other Linkin Park fans!

So, I would like to admit this fact and be led in any group that has a test for me. You can prove that you're a real hardcore Linkin Park fan because you went to a dozen concerts, and you got the signature on your wallet that you went to these concerts, and only the people at this concert could have this. And therefore, you can be led in.

It really clicked when I realized that building a decentralized dating app is not possible or extremely hard.

You have this issue of discoverability and privacy.

And it's not privacy in the sense that the government shouldn't know what I'm doing. No, it's more that I don't want to reveal my dating preferences publicly, for example. And you have no decentralized way of doing this.

I'm interested in the ability of cryptographic tools to allow better discovery on the internet and to create experiences on the internet that really mirror the way you meet people in real life.

So, somewhere between insulting people on Twitter and getting to know them in a group chat, there should be layers of decentralization, discoverability, and privacy preservation.

Who to watch in trust infrastructure

I really encourage everyone to look at the work that Cursive and Tonk have done. They are at the forefront of using tools like MPC and others for discoverability to solve what they call the big problem of the internet: How do you find people on the internet? This is a weird thing to say, but it's still very difficult to do.

I also recommend a great article called On Trust Infrastructure by Arnaud Schenk and You Should Build Trust Infrastructure by the Tonk team.

One of the things that really spoke to me was the idea that we could use blockchain to create more truth. That was really cool. I would like to have more tools to trust people: trustlessness is cool, but trust is also cool.

Sylve was again invited to the House of ZK podcast to chat with Alice Liu, education lead at ZKM, and Ming Guo, chief scientist at ZKM. They discussed trustlessness, borders on the Internet, and what ZK brings to our shared cultures and relationships.

Watch the video or read our highlights below!

Blockchain and allowing people to create truth

When you start looking at something like Bitcoin, you can say, "These transactions are immutable." A transaction that goes through is done, and there's no way to go back in time. And then you start questioning: it doesn't seem that there are a lot of things in the world that have this property of what Josh Stark calls **hardness,** things that don't change. It has this property of immutability that is just as trustworthy and eternal as gravity.

I think if I hadn't gone into technology, I’d have practiced law. The fun thing with the law is that it's absolutely everywhere. Every single interaction that you have is regulated, and there is a contract for it, and everything is mediated through that. The fun thing that I realized in my law classes is that the law is very monopolistic, and that has not always been the case.

In Europe, in particular, we love to think about the Middle Ages as a horrible time when everyone was under a sovereign king, but it was a very decentralized time.

One of my favorite pieces of trivia from the 13th or 14th century was that there were many competing standards for law and also competing standards for coinage. So, people were selling their money, and because they were trusted merchants, they would go around different fairs and villages and sell their coins. And because they were reputable merchants, you’d trust their coinage. You trust that they actually put the right amount of silver or gold inside their coins.

I've always been very interested in how you allow people to create norms and how you allow people to contractualize. The first time I truly realized what blockchain could give us was, oddly enough, NFTs.

The art market is regulated by intellectual property law. In order for you to sell something, it needs to be represented as something that you can commercialize within the realm of IP law or in the art market. And when I studied the art market, I was looking at an art piece, which was a video. And I thought, okay, that's actually pretty cool. It was a looped gif of a guy going up a pool and then jumping into the pool.

And I was wondering: “How do I buy that? How does that work?”

I realized that in order to buy it, you needed to turn it into an asset. And for that to become an asset, you need a whole installation: a physical TV that would hold it, for instance. You were not able to sell the file. You had no way of doing it.

NFTs tremendously lower the cost of things to be contractualized or commercialized.

And suddenly, you went from, “This thing has to cost like 4K because you need a whole installation and you need to talk to a lawyer,” to “It costs five bucks to go onchain and mint this”. You’ve tremendously lowered the cost of creating norms.

Blockchains are truth-making and norm-making machines, and that's a desirable outcome. Zero-knowledge extends that property even more.

The power of attested data

I recommend reading The Mystery of Capital by economist Hernando de Soto. It analyzes and asks why capitalism isn't always the solution. In some countries where you had sweeping liberalization actions in South America, it didn't lift poverty. But capitalism is a great tool: you can use capital, you can have a house, and you can use it as collateral to get credit. Then, you can buy new things and start a business. Why didn't that work?

He stated that some things were not considered within the realm of commerce. So he went around a favela in South America, and he asked, " Who does this house belong to?”

He realized that there are informal ways of knowing how much a house is worth and who owns it. This works because it's local. The 200 people living in that neighborhood all agree that that person owns that house. But if you go a mile away, people won’t know that. And if I look at the government records, nothing tells me whose house it is, and therefore, I cannot trust that.

The book said that if you had a way to represent these intersubjectivities, like these networks of truth, link them with each other, and make sure that they can be integrated into something bigger, you could have access to more capital. And this could be very useful. But this is currently dead capital, things that cannot be used outside of their own little realm. That's the great thing about the dollar, for example. It’s a fantastic API for value: people are always going to accept the dollar. It's much more difficult to use your favela house as collateral.

One way that we solve this problem of intersubjectivity is to put everything onchain. If everything runs on the blockchain, then there's no problem because everything is laid bare. And it's transparent, and therefore, you can create DeFi about it.

But the past 15 years of blockchains have shown us that we can’t put everything onchain. It's not performant enough. And some things were never bridged. There are still things that are going to stay in the real world.

What I really love about ZK is that it can use attested data to produce statements about them. And so if 200 people decide to trust that this house belonged to someone, they can collectively sign this fact. And if you trust that group, you're able to extend that fact somewhere else. Like, I have proof that this person has this house, attested by this group of people. And you're able to have things like transitive trust: if I trust you 80% and you trust that person 80%, then I will trust them by the factor of 80% of 80%.

And that's really cool!

That's why I talk about 50 shades of trust, having this ability to have things that are not truthful and not trustless. You've got something in between, and that's really neat.

Shifting to a different truth model

I remember an old The Economist article saying that the worst fate for a country is not to be like Greece and fail overnight, but rather to be like Argentina, one of the most powerful countries and best places to live on earth at the beginning of the 20th century and then brutally collapsing because of hyperinflation and cronyism.

I like blockchain because it's an engine for pluralism. This means that you can use different methods and speed up the rate of experimentation with currency.

Whenever I talked about cryptocurrencies to my friends about 10 years ago, they would always answer, « Why do I need this? I have PayPal. » And I replied, «You have Paypal because of 400 years of legal history; you have banks, and you have IT systems that make it possible. What blockchain allows you to do is rebuild this whole thing in a few thousand lines of code.»

And that is incredible! If you're able to speed-run 400 years of commerce to get to the same point, what else can you build on top of that?

That has led some people to think that we've solved the problem. They think we're just going to have this machine that is never going to change; there will be 21 million Bitcoins, and everything should run on Bitcoin.

That's not a good idea.

I think that's also very crude intellectually: things are more complex than “no inflation, no problem.”

Bitcoin was born out of the cypherpunk movement. There have always been a lot of libertarian thoughts in there, and I also felt that way when I was 15.

But then I grew up and realized that it's great that there's democratic oversight over a currency. It is great that, in some instances, you need to devalue. It is great that sometimes you can change the flow of money.

Obviously, it's a tool, and like any tool, it can be used in bad ways.

But the result was that in 2008, we were so scared of this lever being used the wrong way that we just decided to remove it altogether. And so we ended up with trustlessness. We thought we had solved all the issues, so we created this tiny little island in the middle of the Pacific where everything was well and dandy. And we thought that if everyone moved over to this tiny little island in the middle of the Pacific, everything would be great.

The past 15 years have shown us that no one wants to go live on a tiny little island in the Pacific.

People want to live where they belong − in Berlin, in Hong Kong, wherever they are − and use the currencies that they have. That's why there are so many debates about whether people should use ETH or native cryptocurrencies rather than stablecoins.

There was a standard called ERC20C, proposed by a team at Stanford about one or two years ago. The idea was to allow token transfers that could roll back, which is a great feature. I would give this to my parents! I would even have this for myself. I would be very happy to have a way to call someone and say, «Hi, sorry, I've made a mistake. I want to roll it back. »

They got lambasted. They were told that this is not what blockchain should be about. Blockchain should be about pure, absolute trustlessness. And if you make a mistake, that's your problem, which is a statement that people usually have until they make a mistake and send $2,000 to a random address, and they can never get back. It is great to have these fail-safes in place.

And that's what's great about cryptocurrency.

You have the option of trustlessness, but I strongly believe that the baseline shouldn't necessarily be that. It should be an option; that’s it. Not everyone is going to want to wear the tinfoil hat. If you're very happy with the hardware wallet and the 24 words buried in your backyard, good for you. But most people are going to want to use banks.

ZK can interact with the real world

The core problem of blockchain is that it's a great norm-making machine, but it's a very short-sighted machine that can only understand whatever the machine understands.

Whatever is on the tiny island in the Pacific can be reasoned about, can be computed over, and that's fantastic.

But that only works for blockchain-native things.

How do we make sense of this with the rest of the value and the rest of the information that is outside of the blockchain? It's a very isolated environment that understands itself. But how does this relate to the rest of the world?

So you end up with bridges and oracles, and it becomes a nightmare.

ZK really shines in making sure that it does not add more trust assumptions to the attestation. So, if I have a passport and I am able to produce zero-knowledge proof that I'm a French citizen, that's a cryptographic statement. The proof does not add more trust assumptions to that. And that's something I can directly put onchain. That's something that the blockchain machine is able to parse, understand, and use without adding any other trust assumption. That's a very good thing.

It's directly from attestation to the blockchain machine, and that's really, really useful. That's the most useful thing that we have for ZK.

Obviously, if you have a shit attestation, it is going to be a shit proof. The proof itself is not going to change anything. If the government's key leaks and you end up with a fake passport, you will be able to create fake proofs. That's a given. But that's a good thing. It means there are no additional trust assumptions: garbage in, garbage out. But if you trust the original attestation, then you're good to go.

Challenges of going onchain

One thing that I've realized with ZK is that it makes it very easy to create boundaries.

For example, at DevCon, you could get a discount if you could prove that you were from a Southeast Asian country. So you would generate proof from your passport that you're from a Southeast Asian country and you're able to get in. It's a way to delineate a circle and decide who's in and who's out. And that's for a really good reason. And honestly, that was a really perfect example: to do otherwise, you would need costly KYC.

Zero-knowledge reduces the cost of discrimination. And when you start looking at it this way, that doesn't sound so great. Does it mean I'm able to discriminate against people more easily? Is the marginal cost of racist acts going to zero?

ZK extends attestations and signatures. If we're going into a world where you start to sign more things, it means that you can extend them, reason over them, and commercialize them.

For example, in France, we have a signature standard called 2D doc, which signs many official documents. So your tax income reports and utility bills are signed, which is really good. You can attest to their veracity, which is really useful.

But then you can get into a situation where, in order to get to an online service in France, they're going to be geofenced. And you will constantly have to prove that you can be let into the inside group.

We have the technology to make that more palatable or easier. But if we start signing more things, we may end up in a situation where you're just going to have to constantly prove and show something that makes it so that you can be let in the circle.

This ability to delineate and create boundaries is fantastic. It's really great. It's a great tool. But at the same time, when you really lay it out, it's a way to discriminate more easily. That's pretty fucked.

Discoverability and borders on the Internet

Here’s a line of reasoning that I get from both Arnaud Schenk from the Aztec team and the Tonk.gg team, Baz and Goblin Oats.

I reread John Perry Barlow's Declaration of Independence of Cyberspace, which was written in the 1980s in Davos.

What he says is very telling: this guy in Davos declares that we're independent, like a completely different country, and that laws do not bind us. He also uses a metaphor where we have no bodies. We exist in this ether, but unintended of the internet, and there, we don't have a body. There's like one global consciousness.

There's a whole body of research about the fact that a lot of the early internet thoughts were really embedded in this new-age idea of being connected with this giant World Identity. It’s obvious when you read that text that they really believed it would erase all the differences between all the people.

They thought that if you could speak with someone on the Internet, connect, and share information, then all the problems would be solved.

Forty years of the Internet later, you realize that if you put people on the Internet, they argue and send slurs at each other, which is pretty bad.

It's weird to say that this was a completely unforeseen event. We never thought that helping people connect with no sense of physicality means that there are no customs.

Discoverability on the Internet

There isn't much discoverability on the Internet.

You're either completely anonymous, or you're posting your entire life on Instagram. It isn't easy to have layers of interaction with someone.

Something that I love about the Tonk team is that they're looking at ZK and cryptography as a way to mediate experiences between people on the internet the same way you meet people in real life.

If you go into a bar, you can't hear everyone's conversations.

You can see vaguely in the background that there's maybe two people on a date, then there's a group of four people over there, and then you start talking to maybe one person, and they reveal some things about themselves, and they're like, you know what? Come with my buddies; I'm going to buy you a beer. And then you get led into this other group.

If you take this metaphor for how the internet works, the first possibility is zero discoverability, which is group chats. For example, I don’t know where the ZKM Slack server is or how it can be let in. There's no discoverability for me about this.

On the other hand, I can go on Reddit or Twitter and then have direct access to 6 billion people simultaneously.

But there's not really this idea of « I could let these people into my group » or « I could find this group with these people » because sometimes you don't even know the group exists. There are layers of interaction between people, which comes back to the thing we discussed about boundaries. I think it's interesting.

I’ve been wondering whether there are alternatives. An example of this is alternative, decentralized designs for Twitter, like Farcaster and BlueSky. But these alternatives are predicated on the fact that everything is public at all times.

It's much more difficult to build a decentralized dating app.

The reason is that you have data that has to be private or data that has to be revealed to a specific group of people. Because I do not want the world to know that I'm a Linkin Park fan, but I would really love to meet other Linkin Park fans!

So, I would like to admit this fact and be led in any group that has a test for me. You can prove that you're a real hardcore Linkin Park fan because you went to a dozen concerts, and you got the signature on your wallet that you went to these concerts, and only the people at this concert could have this. And therefore, you can be led in.

It really clicked when I realized that building a decentralized dating app is not possible or extremely hard.

You have this issue of discoverability and privacy.

And it's not privacy in the sense that the government shouldn't know what I'm doing. No, it's more that I don't want to reveal my dating preferences publicly, for example. And you have no decentralized way of doing this.

I'm interested in the ability of cryptographic tools to allow better discovery on the internet and to create experiences on the internet that really mirror the way you meet people in real life.

So, somewhere between insulting people on Twitter and getting to know them in a group chat, there should be layers of decentralization, discoverability, and privacy preservation.

Who to watch in trust infrastructure

I really encourage everyone to look at the work that Cursive and Tonk have done. They are at the forefront of using tools like MPC and others for discoverability to solve what they call the big problem of the internet: How do you find people on the internet? This is a weird thing to say, but it's still very difficult to do.

I also recommend a great article called On Trust Infrastructure by Arnaud Schenk and You Should Build Trust Infrastructure by the Tonk team.

One of the things that really spoke to me was the idea that we could use blockchain to create more truth. That was really cool. I would like to have more tools to trust people: trustlessness is cool, but trust is also cool.