Reimagining money with zero-knowledge proofs - Jason Morton

Transcript

Um, not actually a super technical presentation, >> [music] >> but it uses a little bit of those all those pieces of psychology, economics, and and mathematics. Um, and not even uh uh fixing, but perhaps replacing replacing money. Um, so let me begin with a couple of sort of data points. I think some people feel with sort of meme stocks, with meme tokens, and a bunch of other factors that um kind of financial nihilism has has been on a rise in recent years. Um, some of that is is arguably just cyclic, like, you know, Dalio would argue that we're a civilization that's based on a certain reserve currency and a certain debt

system, and then there's this natural debt cycle. We're reaching the end of that cycle, and that causes some some interesting effects to happen, which are not the first time that's happened. Um, as that decays, and you know, this guy uh who has a lot of money thinks money will go away or won't matter few years somehow magically, um because AI eliminates scarcity. Um, and so and there's a sense that this thing, money, and maybe more broadly, systems by which we have a bunch of tokens or a bunch of shares, they have a price, the last price they are traded at, and we multiply those two numbers, and we use that to organize all of our activity, that that system may have

reached the end of its useful life. Um, and so what I want to talk about is why I think that's true, how to put it in kind of a little bit historical context, and what might come next. Um, and in my opinion, it's not just another fiat. It's unfortunately not ETH, it's not stablecoins or UBI, or everything is free. It's something that we haven't seen before. Um, okay. So, one way I like to think about this is taking what you might call a Martian anthropologist view of how our economy works, right? So, and the point of that is is not to satisfy this Martian Martian anthropologist, but to kind of force us to step back from the society and the

system in which we have been embedded since we were born, and just think for a second well hey, if I really look at it from, you know, a million feet view, it it's kind of weird and it doesn't make that much sense. And, you know, say for example, you do something for a friend. You You move a couple of boxes. Um you do something. You give them advice. Um you give them some sort of favor. Uh sometimes there there's an ongoing relationship between you and that person, a family member, a tribal member, you know, part of a tight-knit friend group. There's no There's no record. You just kind of keep track of it, right? Um and sometimes, especially if there's no relationship, there is a transaction. Both parties write it down, um or at least uh the

favor receiver gives a piece of paper or a database entry is modified in favor of the favor um uh giver. And sometimes there's both. There's a relationship and there's still a transaction. And the Martian anthropologist looks at this and says, "Okay, what are these crazy humans doing? I don't get how they decide what is just a normal cooperation activity between friends or or or a tight community and what counts as a transaction. It seems really arbitrary to me. How do I know that that, you know, moving the box um results in an exchange of currency?

Um and so the way I want to explore this is through something I call a book of deeds. It's kind of a parable for thinking about, you know, a priori, how would we make solve this kind of problem? Um and you can sort of think about a book of deeds as as a bunch of receipts. Um everything I've ever done, but with no prices attached, right? So, every person has their own book of deeds. Um and in this book, the favor giver and And favor receiver, you know, the the uh the I I bought a coffee. I I gave you a coffee or you know, you moved a box for me. We record every favor we've done.

There's no price attached. We just we just note that it was done, right? And then when we decide as a society or as a new interaction, should we give you a new favor? You know, maybe you ask for a sandwich. I say, "Sure, I'll give you a sandwich." You ask for a yacht. I say, "Uh you maybe not." We look at the book and we ask ourselves, do all the things that you've given others outweigh all the things that you've taken, right? Is your sort of social credit balance positive? Okay, so that's not scalable even for one person. This book gets really big over the course of a lifetime.

Um but and I would argue that our economic systems both the kind of tight-knit group economic system and the big scalable money-based economic systems are basically solutions to this issue that the the book of deeds doesn't work at scale. And they take a very different philosophy. Okay, so this book of deeds, we check whether things that someone has done for others outweigh the things they haven't uh they've done for them enough by enough, then we give the favor and record it, and we've basically developed two strategies as human groups for uh making this book thing work. Let's call one of these forager mode.

That's the tight-knit family tribe mode. You know, most of human history we were we were foragers. 99% or 95% of of an anatomically modern human Um this book that we're using is in this case only approximate. Um it's tracked among the close group. We just keep it in our memory. Oh, so-and-so is a good citizen of the group. You know, we should share food with them. So-and-so is a leech and always takes food and never shares. You know, when he kills a mammoth, he just throws away the rest. Um you know, let's kick him out, right?

Um it's complex, it's fuzzy, and it kind of embraces the complexity of humans. It's In my In my opinion, it's kind of there's a dignity to it, right? Um and then we have what I'd call the farmer mode that only emerged with agriculture. The book is formalized. Um you know, we actually write down every deed and we write a price next to it. Um we have notion of debt that has a specific value. We use money. We use tokens to track ownership of a community. Um there's balances and the whole thing is beautifully scalable. You know, we scale it to 8 billion people, but it's arguably less human.

And it's actually really new for us as as kind of evolutionary entities. It's only a few thousand years old. Um So, I want to zoom in on both of these. One of them um is this first idea of forager mode. You can think of it as like humans as social computers. Um informal exchanges of favors uh tracked in human memory. So, it works really well in small groups, families, tribes. You know, a good early startup or great group or organization feels this way. Um often times it's less than 20 people. More generally, it's maybe like a Dunbar number, 150 to 200 people.

I like the example of barn raising in in kind of the middle Midwest America in the 18th and 19th century. You're part of this group of people. We're all sort of on the frontier together. I can ask my neighbors to help my family raise a barn. Um build a barn. I can ask for one barn, but if I say, "Hey, could you give me 100 barns?" They're going to say, "No way." Right? So, like that ability to call on my community to help, in this case, build a building is a critical part of my family's wealth, but it is not in any sense tokenized.

It's not like I have 300 shares of, you know, Grinnell, Iowa and I can spend that on a barn, right? It's a much more informal, although critically important aspect of my wealth. And when it works, arguably, it feels really good. It's fun. I feel loved and love, you know, back to my community. Um and so there's kind of a question, you know, that we we've got this scalable thing, money. Is it sort of a dystopian version of this, quantifying the unquantifiable value of a human to his community? Um is it scalable but in some sense unnatural? Is it myopic? Does it only see a tiny part uh imperfectly of kind of what we are and what we're doing?

Um I heard a talk yesterday by uh Marian um uh from the CEO of Burning Man. Um there are hints of scalability of forager mode, right? So that has kind of an interpersonal culture based on an economy that's a non-transactional economy, gift only. It can hold together for a week. Um and I'd argue that humans are always kind of balanced on the edge of these forager and farmer mode, and it takes us sort of hours to days to switch modes. And so people show up in an an event like that and after a day or two, they really become in the spirit of it. And the way that their interpersonal activities work, you know, how easy it is to talk to another person, the high trust barriers, everything works. And, you

know, they say welcome home to each other because in some I would say maybe because um it does feel a bit like home in the sense that it's longer of our our evolutionary history. Um and it has, you know, ambitions for scalability to be a culture uh and not just an event. Um okay, so now let's talk about farmer mode for a second. And you know, my one of my thesis here is that farmer mode is really an early paper and pencil version of zero-knowledge proofs, which is why it has anything to do with this session on zero-knowledge proofs.

Um so, you know, how did farmers solve this book of deeds scalability problem? They had a couple of big innovations. One of them is accounting, right? Like actual record keeping, single entry, double entry accounting. This is a really early example of a company that was founded with 15 kg of gold 4,000 years ago. Um and they recorded, you know, what were the terms and, you know, what was the bond curve if you wanted to leave early and all this stuff. Um, this guy is Friar uh Luca Bartolomeo de Pacioli. He's the creator of double entry accounting. He's a friar, so he was a citizen of of an early network state called called the church. Um, he lived from 1445 to 1517. And um,

this technology of accounting, incredibly precise, incredibly scalable, number of units times the price of that unit. We're working with pen paper, so one or a stone, so once you write down a price, we can't change it later. Right? It's it's it's not a computer screen. It's fixed in a book, literal book, or fixed in stone. And so prices can't change that often and they certainly can't change um, you know, once time has passed. A second huge innovation of farmer mode is fungibility, right? So, we have multiple groups who each keep these records and maybe they have a banking system that lets those records be shared

among a medium trust community like a country, uh, what we would now call a country. Um, and people figured out that it's useful to have some kind of standardized item that we can use to trade across boundaries between hostile tribes, between people who don't have a an aren't part of the same network of of accounting. Um, and so about an hour and a half from here is this town of well, it used to be called Joachimstal. Um, and they had a really good silver mine and in that silver mine they started printing Joachimstalers, which were uh, very precise amounts of silver.

Um, and this idea of like the Joachimstaler always has just exactly the right amount of silver became a standard and from this coin um and that's that that fungibility of that coin is where we get, you know, the word dollar which is now a very popular um currency and it has that same same transferability, right? So, that was I mean, many people have used specie but this is this is actually really important one that's that's that's from um a mine just just just an hour and a half drive from here. Um so, fungibility between groups, we need to be able to transfer that wealth that's accounted for um between larger civil larger chunks of civilization.

And so, bringing these two things together, we get the idea of compression. So, money itself is a zero-knowledge proof. It's a compression technology, right? So, it's a way to get these book of deeds as summarized by prices, crunch them down to a single number. So, when now when I want to want to ask for a favor, you know, I want to go get a coffee from the stand a couple of rooms over, it the society really does have to balance, you know, all the things that I've done for others, all the things that others have done for me and the way they do that is they call up my bank and they say, "Oh, you know, does this person have enough credit on their card to buy a coffee, right?" And that allows me to prove to the coffee

seller to ask if coffee um that my balance is positive, that I've, you know, if you look at all of the history of all the things that I've ever done and have done for me, that nets to a positive number. I can prove that to them without sharing uh with the barista, you know, let's go through my book and talk about, you know, the things that I did for John in second grade, right? Um so, it's a compression technology. I can provide this without having to reveal and tally everything every time. Um and we outsource the management of this proof system to a network of banks, typically banks and other financial institutions.

They keep this book of deeds. They you know, really that all these receipts do hit, right? Those records exist somewhere and they aren't thrown away. Um and they release only the summary, right? So, they release just the the tiniest bit of information that you need, yes or no, do I get to buy the coffee? Do I get to buy the yacht, right? Um so, it is a zero-knowledge proof of the balance of favors that I have done for others. I prove that to someone else. And it's fungible um because of the Yocum Stoller and its descendants, it's fungible. So, I can take that uh proof. I give you the proof that I did this thing and then you can take that proof and recursively prove it to someone else, right? I've given you a dollar and then you can take that dollar and and and pass it to a friend um or

someone you want to transact with and and that uh they accept that, right? So, in same sense as as zero-knowledge proofs, there's a recursion to it. Um great. So, money is a zero-knowledge proof. It's proving just one number. That's kind of the the takeaway. And that was really a computationally efficient. We were able to scale to 8 million people with pen and paper and then later computers doing the same things. There's a sense that more and more of human interaction is getting financialized and transactional. There are good aspects of that, but I think overall it feels kind of bad.

Human actions, values, our contributions uh and the way that we interact with each other are kind of reduced to a number. Um and hey, it was it was good job. It was amazing given the trade-offs. But, you know, times have changed, right? We have new technology. We have the technology underlying blockchain. So, we have consensus algorithms. We have cryptography. We have programmable cryptography. We have zero-knowledge proofs and we have AI. Um the trade-offs have changed. The Pareto frontier has shifted. And I would argue that the way that we've used cryptocurrency so far with tokens uh is kind of a skeuomorphic porting of the same farmer model that was first written done in stone and then in paper and then with databases to a new technological substrate. Uh but

we're not restricted to that. There's no reason why we need that economy to work exactly the same as the old economy because the computational constraints have been lifted. So, my hope is that crypto is actually just the bootloader for whatever comes next in our economic system. Um we have incredible new primitives that that make new things possible. We have an opportunity to kind of maximize human dignity as our objective function for whatever we create. Don't reduce to a number. Uh I'm not going to have an answer for that how to do that, but I want to observe that a lot of experiments are possible and scalable that were not possible before because of the technological changes. We can let communities decide what they value and

how that changes with time. Um and you know, maybe it's possible to scale the mode, the forager mode, that feels good and human in a way that never was. This isn't capital communism, it's not capitalism, it's a secret third thing that was not technically feasible ever before in human history, but it is feasible here in 2026. Okay. So, the last thing I want to talk about, again, I don't have the answer for this. Uh I want to encourage people to do experiments and figure out how outfits could relate to each other. So, I'm going to talk about a couple of experiments that I think are fun that we could do.

Thought experiments, really. Okay. Before I do that, I want to talk about one a very opinionated desiderata that I think factors into everything, which is that I feel like the direction of bearer assets is not the right direction. Even with privacy, that makes it worse. Um cuz a bearer asset is kind of a book of deeds that's owned by whoever picks it up, right? Like I can have my book of deeds that goes back to first grade and I accidentally leave it and if somebody picks it up, it becomes their book of deeds. And that's not a good system. Um and therefore, you know, I think community-based kind of whatever comes next requires identity, not necessarily in the legal sense, passport, but in the sense at least of a costly persona. You can have more than one identity, um but you you should the identity itself

should not be transferable. Um the other problem with bearer assets, of course, is that a financial system um that has all the things we love, like decentralization and uh bearer properties and censorship resistance, uh and programmability, depends on flawless software, depends that the humans and the AIs never make an error in coding, um but everything has errors. You know, if if somebody at JP Morgan makes a coding error in their banking software, all the customers of JP Morgan don't lose their money. Um I think, unfortunately, that seems to be where we're stuck uh in the crypto world um in DeFi and everything else, and I think we need to figure out a way to to fix that reliability in order to be

widely useful and have a have a broader system. So, so bearer assets, of course, have limited utility as a foundation, um but possibly that is mostly about uh symbolic value and to be used to trade fungibly with hostile tribes. Um so, they always have a place, but they they can't be the day-to-day system. Um okay, so here are some examples. One is the AI command economy. Okay, we keep the book of deeds. It used to be not scalable, but now it's fine. We keep these records anyway. The record of everything I've ever bought and sold exists in computer somewhere.

We take that whole book of deeds instead of passing it through a bank and doing a one-parameter latent vector version of my social credit balance to decide whether I get to buy a yacht or buy get to buy a mango, we just throw it in an LLM, and we say, "Hey Claude, you know, hey hey chat GPT, should I be allowed to get this mango? Okay. So, this is kind of funny and it's a little bit dystopian, but it's interesting the point I'm trying to make is it's interesting that it's actually possible, right? So, like that was never possible. It's totally possible now, right? We don't have to attach prices to things.

We could simply let an AI decide, do we get the next item or not? Um another version of this is kind of vector money with retroactive prices. So, here's taking the argument that, you know, the fact that we the amount that I paid for a gallon of milk or a gallon of gas in, you know, 2005 was an optimization that was well suited to stone and sort of well suited to paper and not really necessary for computers. Um so, we keep the book of deeds. Maybe we reduce the dimension to some factors like, I don't know, energy price or, you know, carbon or something.

And then the historical prices of things that I've done float. They change with time. So, as we gather information on their positive and negative externalities, so for example, um maybe I bought a gallon of gas, it seems like it should cost a dollar when I bought it, but later we realize that global warming sucks and we wish we hadn't done that and so we change the the retroactively change the price. We say, "Oh, those gallons of gas are now $3 when we make a decision as to what your contribution is." And that sort of changes everyone's time horizon for the actions they take and makes everyone a little bit predictive of what they're doing, right?

There's a little bit of flavor of something like retroactive public goods funding. Or maybe you took an action to you know, you were a teacher or you helped a child that's in a way that seemed like a minor kindness, but she ended up changing the world and the help that you provided to that person is retroactively repriced. So, that's that's something that we could do now with computers. Again, not necessarily saying this is what we should do, but it's it's a thought experiment to say it's interesting that we can do it. Um another one is a similar but with community values. So, you keep this dimension reduced book of deeds and every community, whether it's a geographical community, a country, or a network state, weighs those deeds differently according to its

values. In a more eco-conscious community, that gallon of gas may be all gallons you've ever burned are $10 a gallon um this year or, you know, permanently. It costs your reputation more. Actions that you taken in the past because this community has different values around actions. Maybe in a health-conscious community, all the healthy food that you've bought is cheaper, not just now, but retroactively in terms of their calculation of where you stand. Um, so it's kind of the your money is no good here um discounts and stuff made sort of more precise and specific.

Um, and then there's one that I call fuzzy money, which is um I is a complex idea and it's a whole mathematical um construction that I have and I won't get into the full details of, but the idea is basically instead of thinking about token count or share count times last price as your value, which doesn't really reflect how much value you put into a system or how much you took out. It's precise, but it's not accurate. You, instead of saying, you know, there's this pool, a company say, and you own, you know, 1.256% and you own 3%. Um, instead you use a distribution over possible ownerships using the Dirichlet uh distribution.

And then every one member of the community has this kind of shifting and fuzzy claim on the resources of the community, like people with the barn raising, um but it's tracked with computers instead of relying on everyone's minds to remember it. Uh, and what's neat about this is that it gives you new macroeconomic parameters. So, uh the inequality in the society is just a dial you can shift. You can change how peaked or or flat these distributions are and the definiteness, like how tight or how spread out that distribution is also is another parameter you can change. So, you if you feel that, you know, things have become too unequal, you can spread them out. If you feel that incentives aren't precise and clear enough, you can

you can tighten it up and make it look to behave more like money. Um So, another example of some crazy stuff that we can do, completely feasible now, a bit impossible with with tablets. So, in summary, you know, these things are just thought experiments. I think money was cool. It got us pretty far, but it's an obsolete technology. Uh so, I invite you to do better. The design space is wide open. Thank you.

>> [music]