Radical Crypto Imaginaries

Well, I can only hope to live up to that introduction. Good afternoon, everyone. Thank you for coming back from lunch. I'm flattered that you picked me over finishing your lunch. It means a lot.

So I have a few caveats to begin with. One, there was a title on the agenda. I think it was called Radical Crypto Imaginaries. That is a great title. I didn't write it.

Someone else wrote it. If you're here for a talk that was titled that, then sorry, because I wrote a different talk, whether that title will fit or not is up to you to judge at the end. Caveat two, this talk is not about Enoma, really at all. Caveat three, this talk is not about privacy. So if you were here for Web3, at least not really, if you were here for Web3 privacy stage because it has the word privacy, then just caveat emptor, I told you up front, you can't blame me afterwards.

So speculative cryptosynthesis, how to maybe turn around the Titanic. I want to start this with an anecdote and a question. The anecdote is when people ask you, what do you do? Where do you work when they aren't crypto people? What do you tell them?

Because when I tell them, I tell that, you know, I live here in Berlin, I tell them I work in software and software development. Right. Not exactly a lie, but also not exactly the truth. And I think this brings us to a central question. What is the industry?

Two answers, random answers. Who has a good definition? What's the crypto industry? Legalized theft. That's one definition.

Any other definitions? Decentralized money printing. Well, you don't even need to be here for this talk, man. OK, so I think it's nice to look at other industries that exist in the real world and do stuff for a long time and have developed some kind of clear structure that allows us to look at them and use them as models for comparison. And one of my favorites is automobiles, because everyone's, you know, ridden around in a car.

It's pretty tangible. So just very basic, you know, sky high, no data picture of the automobile industry. What does it consume? What does it produce? What are the conditions of its sustainability?

Well, the automobile industry consumes material resources. You need some stuff. You need some rare metals. You need some plastics. You need some less rare metals to make this car.

It consumes some skilled labor. That's kind of what defines the industry in terms of an actual practice of production. There are people who are good at making cars or at least they're trying to be good at making cars. Sometimes they move between companies. They write their practices down.

Maybe they try to trademark them or patent them. Maybe not. But eventually there's some kind of knowledge sharing. There's some shared practice of production that gets built and at least we would hope gets better over time. And while the automobile industry produces cars and you as a consumer or just someone who's paying can judge whether or not the cars it produces are successful by whether or not you can drive them.

Does it drive around? Does it feel safe? Does it go enough miles without needing more gas or electricity or whatever? So I would argue in order to sustain itself, the automobile industry needs to keep selling cars, right? That's what people buy.

That's what pays for your job. If you work for Ford or Tesla or BYD or whatever the new car manufacturers are now, sometimes they seem to be also making cell phones, Xiaomi. We can reduce this question of, quote, is the automobile industry sustainable to something like, do people want new cars? Right. And maybe they want them because you've sold them.

Maybe you've done some preference manufacturing. But, you know, still, at least there's a car, right? Either the car drives or it doesn't. There's this kind of connection point with physical reality, no matter how good Elon Musk is at marketing. If the Tesla runs out of battery in 50 miles, it's just not very useful.

Right. So there's a clear picture of what makes this industry sustainable. Do people want new cars? What about crypto? What does crypto as an industry consume and what does it produce?

Well, about seven years ago, I decided to join crypto. And at that point in time, I thought that crypto as an industry was like a technology industry. I thought that it sort of consumed skilled labor. There was some shared practice of database cryptography production. It was open source.

There was a lovely culture around knowledge sharing. There was some value placed on privacy, on self-sovereignty, on decentralization, on autonomy, on just legibility of what users are doing with these kinds of large distributed systems. And, you know, I thought crypto produced them. I thought crypto produced distributed databases. And I guess I don't know what people do with them.

They send transactions like it's a little abstract, but at least there's a database. You know, there's some software on GitHub. You can look at it. You can analyze it. You can see what it does, whether or not it does what it says in the box.

And I thought, you know, in this world, the question of is the crypto industry sustainable could be reduced to the question of do people want new and better distributed databases? I thought the production function of crypto was something like can we produce databases that better satisfy what it is that people want from them? Right. That's why I started long ago on the Sonoma project, for example. So over the years, I've come to be at least come to the belief that I was pretty naive in this conception of crypto as an industry.

And I think an easy way to evidence this is just to look at the actual, you know, look at where the money flows. Look at the top 11 projects on CoinMarketCap, Bitcoin, Ethereum, Tether, XRP, BNB, Solana, USDC, Dogecoin, Tron, Cardano, and Hyperliquid. How much database innovation is happening here? Not a hell of a lot. Something else is being bought and sold, something else is, frankly, funding this conference and it isn't better databases.

So let's try a more realist take. Crypto as an industry take two. Well, it consumes skilled labor. That labor is actually maybe less about building better databases and it's more about marketing things. And what is it marketing?

Well, it's marketing tokens. I forget exactly the line you said, but decentralized money printing. In fact, it's not even very decentralized. There aren't that many big tokens. There's a pretty, pretty steep power law.

So crypto as an industry consumes some shared practice of marketing, of kind of somehow legitimating these tokens. Like, I mean, I can't just launch Kriscoin tomorrow and have it be $100 million. You know, I have to at least pretend to do something. Right. So there's some shared practice of production, but it's not really oriented towards producing better databases.

It's oriented towards something else. And crypto in this in this understanding, this may be maybe even slightly cynical, but I think just more realist understanding. Crypto produces tokens. What can you, the user, do with tokens? Well, you can buy them.

And hopefully many people might think you can sell them to some other some, you know, not even necessarily a greater fool. Like, honestly, you don't care, but hopefully you can just sell them to someone else for more later. Now, in this picture of crypto, as an industry, databases are still involved, but they're not really the primary thing that's being produced. They're there kind of for another reason. They're there for, well, they're kind of there for legitimacy.

You know, how do you sell a token? Well, you kind of have to pretend that it has to do with something, that it has to do with database, that it has to do with a technology platform. The database is still useful, but it's not useful in virtue of being a better database. It's useful because it fits the narrative. So in this, I would argue, more realistic picture of what's going on in crypto.

Mostly the industry is about token production, about somewhat decentralized money printing. And then within that big circle of token production, there's a little circle of database production. And the circle of database production still gets some funding. It still makes some progress. It's still useful, but it isn't an accurate characterization of the industry as a whole.

Follow me so far. Any radical disagreements? Right. So now I want to go back to my earlier question of sustainability. And here, if the crypto industry is really producing and selling tokens, then the question of, is this crypto industry sustainable?

Just like the question of, is the automobile industry sustainable? Could be reduced to, do people want to buy new cars? Can be reduced to the question of, do people want new tokens? And if people are looking for new tokens, would just like them to, well, go up, then probably they don't. Maybe some people can guess which token this price card is for, but it doesn't really matter.

Most of them look like this. And if you are a user judging this on, you know, I'm comparing this to the S&P 500 or I'm comparing this to buying Robinhood stock or I'm comparing this to whatever, not looking so hot. Now, where can we go from here? I don't have, you know, this talk was titled Turn Around the Titanic. And I wrote that title before I actually prepared.

So I got to admit, I don't have a surefire proposal that was too ambitious, but I do have, I wanted to analyze a few paths because I think there's sort of things that are implicitly going on already and it's helpful to call them what they are, sort of name them. And I also think there are possibilities that could be easily, seriously explored that aren't. And at the end, I want to talk about a few of those. So here I'm going to sort of classify path forward. If we take this, oops, come on clicker.

If we take this question, just take it as a hypothesis. You know, I haven't, I've given you some intuitive arguments, not any hard data, but take this as a hypothesis. Crypto isn't sustainable. So what do we do? I see three path forwards.

Two of them are kind of already happening. One, we could become finance 2.0. The finance industry is very successful. It's sustaining itself and it follows a pretty, you know, a pretty simple playbook.

Step one, pay off the imperial officials. Step two, establish a regulatory monopoly. And step three, profit. This is happening already. I'm going to offer two images without comment.

Oh, maybe I'm not. Oh, here's one. Okay. This one you've probably seen. Maybe this one you've also seen.

You know, Coinbase knows what they're doing. So the strategy may work in a narrow financial sense. And for some people, that's what they're here for. But is it what you're here for? I can't answer that question.

Strategy two. This is what's happening, honestly, in Berlin. You know, I'm guilty of this too. I call this strategy, collect your paycheck while the world burns. This strategy involves a few components.

One, come to a lot of Ethereum values conferences like this one, no shade. Two, build some more info, do some more applied cryptography. You know, pretend that you're not really involved in this big meta industry that sponsors U.S. Army parades or whatever.

And I don't know, pray. I just don't find it very compelling. So to slightly misquote Matt Tybee, don't know if I'm pronouncing his name right. Sorry. Who knows this quote?

He was originally talking about Goldman Sachs in 2009. The cryptocurrency industry is a giant vampire squid wrapped around the face of humanity, relentlessly jamming its blood puddle into anything that smells like money. And the problem is the Ethereum values conferences just legitimate this. I like them too. They're fun, but they're legitimating it.

Vampire squid appreciates your credibility. It will suck it dry. So we're here. It's a little depressing. What do we do?

What other options are on the table? I'm just going to talk about three. One, some kind of financial engineering. Crypto is pretty good at financial engineering. It can be used for other purposes.

Some kind of financial engineering for the public good. Could we do that? Two, world coin except like the non dystopian version, which is maybe more plausible than people have thought. And three, vertical integration or money printing bazooka edition. First, financial engineering for the public.

Crypto has just atrocious capital efficiency, like it's really bad. If you look at the total amount of funds that flow in and the total amount of like stuff that gets produced, you know, apart from more tokens, it's not great. But there are lots of examples in history of people doing some kind of financial engineering and marketing that actually generated a lot of surplus. And I'm going to give two here. Who knows the Ansari X Prize?

Right, the Ansari X Prize was offered. Oh, boy, I believe it was in the 90s. It kind of predates me. Is that right? You know, 90s.

Well, Ansari X Prize was designed to lower the risk and cost of going to space by incentivizing the creation of a reliable, reusable, privately financed cruise spaceship that finally made private space travel commercially viable. And it was one years and years later, $10 million in crypto is not a lot of money. You know, more has been spent on like putting ads and cabs in Berlin, not to throw too much shade. But this was very effective. You know, not only was it effective, it's generating the stated outcome, generating progress and commercial space travel, but it was also really effective marketing.

You know, I've heard of it. I have nothing to do with the space industry. You've heard of it. People in this audience have heard of it. Even if what you're optimizing for is marketing, as long as you care about optimizing beyond some little bubble within crypto, this was a good strategy.

Example one, example two, who's heard of the Vesuvius Challenge? The Vesuvius Challenge was a recent effort, I don't know if all the prizes have been claimed yet, but some of them have been, to sort of grant a prize for the creation of machine learning algorithms using, you know, transformers, similar technology to LLMs that could read the carbonized Herculaneum scrolls. Only 1.5 million, you know, really not a lot of money from the perspective of crypto, and it generated results. Someone managed to do it.

Extraordinary. This is technological progress. So three step playbook to actually put this idea into action. One, pick an underfunded real world problem, get out of crypto. Crypto problems aren't real.

You know, go look at something, read, it has to be hard, right? Put a commercial space flight, you know, that read these scrolls that no one has managed to read before. Pick an underfunded real world problem. Create and market a prize for progress. Commit to it up front and pay it out.

Now, what works here is not the money, of course, the money is kind of a small part of it. What works here is that people care about the problems. And someone spent a long time talking about the XPRIZE, they went to conferences, you know, they made social events. They made it something that people really wanted to compete in. In the end, people competing in the XPRIZE competition spent way more than 10 million dollars, right?

It wasn't really profitable. But they were there for the mission. Right, WorldCoin, the non dystopian version. There's my image. So if this guy wants to get his iris scammed, you know, it's his choice.

But for the rest of us, could we make something else? Now, WorldCoin is interesting because it attempts to solve a very real problem that is becoming more real because of one other of Sam Altman's companies, but whatever, you know, create the conditions for your own demand and stuff like that. And that problem is, well, anti-sybil in a digital world. All of the methods the banks currently use are going to break. All of the methods for authenticating people's identity over the Internet that are based on seeming like a convincing human in a phone call or a video interview or whatever are going to break, right?

What do we do? WorldCoin, the non dystopian version. So about six months ago, I was working on some ideas on the research forums. There's a screenshot from this there with some other researchers that I work with. And we designed a concept that we call Chelling Tests.

And the basic idea is, you know, what WorldCoin tries to do is to equate humanity, being human with some physical property, like having a unique iris. They try to equate humanity with like some particular physical thing. And if you have the physical thing, then you're human. And if you don't, then you aren't. You know, I mean, I don't think this is going to work.

Like someone should just make the, you know, unique iris manufacturer and then print WorldCoins, which falsify this hypothesis more quickly than rhetoric can do. But the actual, you know, actual attestation, you know, what makes you human? I don't know. I mean, do I consider you human or not? It's not really an objective question.

So what if we take that intersubjective nature of the question and actually try to make a protocol which reflects it? And this sort of concept we called in our discussion, Shelling Tests. The idea is pretty simple. One, the basis is this kind of semantic attestation. You know, I meet Nick and we look at each other.

Are you human? I mean, honestly, I'm willing to say yes. Right. I'm willing to say yes. And if Nick gives me his public key and he says, this is my public key and it's my unique public key, I'm not giving a different public key to other people.

Will you attest to this? You know, I'll say yes. That's a semantic attestation right there. Now, imagine we have a graph of this, a big graph across, you know, it could just be the people in Berlin or it could be the whole world. We have that graph.

Once we have that graph and all of these, all of these attestations are private. We don't put them on some public ledger, but we have that graph. Now we can perform computations over it. Now I can ask a question like, well, are people, you know, close enough to my trust network willing to attest that Nick is human? Maybe I've never met him in person, but maybe some, you know, three of my friends have, and they've made this kind of attestation.

I don't need to authenticate his identity specifically necessarily, right? But just that this public key corresponds to a unique human. We can create this kind of test on a graph in a way that is not really so dissimilar to how social systems normally work. You know, you ask your friends, it's gossip. Now, luckily, you know, this is just some ideation on the forums, a lot of speculative proposals, not enough actual protocols.

But I discovered just a month after that someone else had already reinvented the thing and done a much better job than I did. Who has ever read A.H. Shapiro? Amazing.

A.H. Shapiro, there are two things that I wish people in crypto would read, like if I had to pick two things, one Shapiro and two Moxie's talk, which I'll get to slightly later in this presentation. A.H.

Shapiro has solved half the problems that crypto has and people just don't read. So specifically, this paper, Genuine Personal Identifiers and Mutual Assurities for Civil Resistant Community Formation, kind of a mouthful, but has invented this kind of protocol. Semantic accessation, some kind of graph tests. It's on archive, it's public access. That would be a thing that people in the world find valuable.

If you could cheat proof of humanity, doesn't require scanning your orb, you know, scanning your eyeball on the orb or whatever can be bootstrapped in local communities, doesn't even require, you know, these are semantic accessations, you're building a web of, you don't need tokens to pay for gas or whatever. That would be valuable. Not even so complicated. Right. I'm running low on time.

So, top three that we could pursue, vertical integration. Who has seen this talk by Moxie Marlin Spike at 33C3? Julio, my man. Yes, the ecosystem is moving. This is the other thing I wish everyone in crypto would watch.

This talk is Moxie Marlin Spike, who was the former founder and chief executive of Signal, defending his decision not to decentralize the Signal protocol. He was being attacked at the time for this by, you know, all the armchair decentralization maxis, not in Ethereum or Bitcoin, but in the federated communications community, Matrix, you know, CCC. Defending his decision not to make Signal decentralized. And the central defense, or federated, I think was the word in use at the time. The central defense that he offered is that the ecosystem is moving.

The only way, Signal didn't care about competing with the other federated chat protocol. They weren't interested in that. They cared about competing with Facebook, competing with WhatsApp, competing with Google Chat. And all of those companies could move really quickly because they controlled the whole stack. And without any kind of control, without any ability to iterate the protocol, to add stickers, to add group chats, to add whatever great features in Signal, most of us probably use today.

Without that ability to move quickly, they would have lost. And you wouldn't be using Signal, you would just be using, you know, Facebook or Google Chat. That was the defense that Moxie offered. And I think that he's basically right. So what would this look like?

Well, this is the money printer bazooka option. One, pick an actual consumer application that just requires a lot of capital and a focused organization to build and market, like payments, I don't know, WeChat, the Ethereum edition, or U.S. Treasuries, but with lower fees, a surprisingly palatable product because so many intermediaries pay so much of the fees. Two, give Vitalik the ability to print each.

I'm not joking. You know, it doesn't have to be an infinite amount, just like one percent of the supply per year can be much less than is paid for staking even. Give him the ability to print it or give someone in charge of the Ethereum Foundation and hold them responsible for actually shipping. Give them the ability to iterate across the whole protocol stack to make the consumer app work. You know, protocols can be still be open source.

You can still bear. But instead of having, you know, five hundred and twenty organizations building small parts of the stack and nothing ever integrated, centralize it. At least the production, if not the operation. I am at time. I hope this is interesting.

Please direct to the email address. Thank you. We will have. Don't go too far away. We'll have time for questions.

They have Nick. Of course. Thanks, Christopher. Yeah, I have reached the same conclusion. It has to be web of trust stuff.

Classic argument is that this doesn't doesn't scale and you get no kind of like nice, comfy, single data point that represents a human or whatever. What are your thoughts on how you break that kind of adoption gap? And yeah, how I agree with you. We have to start going to compete with World Coin. But like how do we credibly sort of do that?

Yeah, that's a good question. I mean, I think web of trust is a broad concept that encapsulates many specific ideas, and there are different specific ideas within that are more or less difficult to actually implement. I think one advantage, both protocol wise and socially and the social part is the much harder part in general. I think one advantage of this kind of just human attestation approach, which is very limited in some sense. You're not trying to build a like variable credit graph.

You're just trying to build attestations of does the public key correspond to you and uniquely correspond to a human or not is that it's easy to understand. Like I can go up to you with my my signal verification, like, you know, UX on my phone app and we can scan our phones. And do I trust you or not? And, you know, if I make semantic attestations that turn out to be false, if you turn out to have been lying, then people can see this. But it's kind of discreet, right?

I don't have to say, do I trust you? Five hundred or a thousand or these kind of arcane questions like no idea. Right. How would I even rate that? But this is a little bit more tangible.

I think another way that it can start is in maybe not even trying to make a global thing, but starting like unique humans, unique public keys within the context of Berlin. You know, you make a phone app to do this. It's pretty easy, especially if it's free. You don't need to pay gas. You can build this kind of network and then local businesses, you know, Berlin coin that wants to offer a UBI for people here or something like this can use this.

It's a of course, everyone wants to distribute their tokens to unique humans. It's actually a real problem. So I think these could be vectors, but I don't have an exact plan worked out, of course. So maybe you're not the person to answer this question, but what's on your name tag? What's on my name tag?

It says I want a circle's invite, but I was somewhat coerced. But I do actually want a circle's invite. Yeah. So circles. Yes.

Basically, it already exists. What you just talked about basically already exists in the form of circles. But it's a payment system. Can we turn it into like an identity project? Right.

Possibly. I mean, one problem with circles, at least as it existed historically, I have not, as this sticker indicates, I have not used the current app yet. One problem with circles historically is that it wasn't private. And if you have a graph of all of these attestations and it's not private, I postulate that actually that might be worse than WorldCoin. You know, WorldCoin is, I mean, run and maybe in a dystopian manner, but at least as far as I know, they're not posting everything on the Internet yet.

So, yeah, in fact, Julio Poetic and lots of lovely folks. It's also a problem we work on in Zenoma are interested in creating the conditions of possibility for something like circles to be done in a private way. Hey, Chris, I was thinking about the earlier comparison you're making between the automobile industry and crypto. Yeah, I think it's a really interesting framing of the question. What does it produce that people actually want?

The I guess the thing that I had in my mind that was tripping me up is that the automobile industry produces commodities like a physical material commodity of the of the automobile. And then I was trying to think of what is the corollary for that for crypto, whether that's like tokens is one potential like fake commodity that is produced. But then I was thinking maybe that is cryptographic proof that it's a production of proof as maybe the thing that the crypto industry makes, but then it's a very it's a lot of different proofs. A lot of things to prove you can do. Right.

All right. I mean, personally. Go ahead. I was just wondering if that fits your. Your comparison or not.

Right, right, right. I mean, I think there are two parts here. One part is this, you know, using the automobile industry as an example. And it's just kind of like what is consumed, what is produced, what is what are the conditions of sustainability lens to look at an industry through? That's part one.

That's like an analytical lens. And then part two is looking at the actual crypto industry as it exists today. And my argument in this talk, I mean, I think proofs could be a valuable thing to produce. You know, they would be a thing that I personally would pay for. However, it's clearly not what is currently sustaining the current crypto industry.

People aren't paying for more proofs when they buy Bitcoin. They're they're paying for you know, they're paying for a token that they hope will go up by and large. This is a generalization. Of course, there are lots of.