It's really nice to do the summit again. The last one was 2019, and it was a great event back then. It's been a great event this year. I really enjoy Funk House, the fact it's kind of slightly more festival atmosphere than most things, and the zero promotional stuff is really refreshing for a convention or conference in this industry. It's a pleasure to be taking the stage for the third time at Web3 Summit.
This year, I think there was some suggestion or expectation I was going to talk a bit about Jam. Jam is proceeding just fine, a lot of development happening, some interesting minor alterations to the protocol coming through that you can track on graypaper.com. But no, this year I want to talk a bit more, I want to continue actually the talk I was doing at Decoded. Who was there for that talk a month or so ago? Okay, not too many, maybe half. Where I was exploring the concept of individuality, personhood on Web3 systems, blockchains, and so forth. I want to continue in that vein today, backing it up with a little bit more concrete stuff. I hope it will be interesting. Something I'm calling Web3 citizenship ideology.
What's it all about? Citizenship was—I haven't bothered looking up citizenship in a dictionary, I don't think it's especially helpful. To me, citizenship is about belonging to a broader collective on the basis of being a person. The biggest point I want to make about citizenship is it's a property that a person has, not a property that a machine has, not a property that an idea has, not a property that an organization has. It's a property of a person, and this is actually the crucial thing when we talk about citizenship.
Citizenship oftentimes comes with rights and responsibilities. If we think about nation-state citizenship, responsibilities are usually some sort of tax thing, maybe a death tax, some potential to be called up to go to war. Rights usually involve, depending on which nation you're a citizen of, some sort of voting right or the ability to have some control over territory. I'd say it's broadly these three things, and it's the first thing I really want to dive into in this talk.
We are fundamentally talking about digital systems. I don't want to make any discussion about non-digital systems. It's Web3, it's digitally native, it's internet native. We're going to take individuality, the concept of personhood, a key element of citizenship, and talk about it in the context of a purely digital system. I want to understand the nature of it because it's only by doing this kind of groundwork that we're going to end up producing anything that's actually truly useful and stands the test of time.
First thing to understand is individuality is not an inherently digital concept. There are some people out there, I'm not going to name names, that have tried to make it an inherently digital concept by arguing things like, well, if you can provide a unique retina, a piece of data that identifies a unique retina, then it can be understood as an inherently digital concept. This totally disregards the concept of trust and that eyeballs are not individuals.
The concept of individuality, in some sense like the concept of citizenship, is kind of fuzzy around the edges. I'll give you a simple example, although maybe a laughable one. Siamese twins, are they individual? Siamese twins that are joined at the hip, otherwise two arms, two legs, two heads, this is two individuals. Probably most people would say yes. Okay, what about one that just has two heads but two arms, two legs, so kind of just one torso, two heads? Is this two people, one person? The reason I'm getting at it is because individuality is not a mathematically definable thing. It's something on which we have an opinion. We ultimately cannot define a truth to it in any useful way, in the same way that we define truth to the outcome of a transfer transaction as being this balance goes down some amount, that balance goes up some amount. That is very clearly a non-arguable thing. Unfortunately, with individuality, we're not starting at this point of mathematics with very clear axiomatic definitions. We're starting at a very fuzzy point, basically one of human projection on the world, and that projection, as with most things human, is imperfect. To try to import this imperfect thing into a fundamentally perfect system is hard.
Like most aspects of the world, therefore, it must be firstly discovered. There has to be an element of discovery, inspection of the world to discover the underlying truth of the world, and then we have to find some sort of consensus on the validity of this discovery. I'm going to dive into this a little bit more, but my main point here is there's no silver bullet. It's not that we can come up with a perfect mathematical algorithm, a perfect piece of cryptography, or a perfect hardware device and believe that we have solved individuality. It's not going to be as simple as that. Individuality is an imperfect concept. To introduce it into a perfect system is going to need some clever engineering and some judgment calls that hopefully we can all agree on, but still make a useful system.
Okay, enlightened discovery. I'm not going to talk about unenlightened discovery. Enlightened discovery, so discovery of things about the world since broadly speaking the Renaissance, the liberal Enlightenment, implies basically three things: evidence collection of the world, so we first collect evidence on the world, we measure things. Second, evaluation of the evidence. And third, judgment of that. These three things are basically crucial to, amongst other things, the scientific method.
How do they map onto the scientific method? Well, evidence collection, I mean, pretty literally what scientists do. Evaluation is essentially the evaluation in the context of a null hypothesis, usually: does the evidence support this null hypothesis or not? Judgment is essentially, is there a broader consensus among the scientific community that the experimentation that has been done implies this particular fact about this particular theory? Does it imply the truth of this particular theory or not? And this, in some sense, is how we as an enlightened society discover facts, truths about the world.
Now, unfortunately, in the last 10 or 15 years, in some elements of society, we seem to have lost sight of this. We see sometimes this very childish nature of personal truth, 1+1 equals 3 because for me I think it does. This is nonsense, this is immature. Unfortunately, it's become a bit of a thing, but it must be disregarded for our purposes trying to build a better world.
Now, Web3 principles imply, as we should know by now, we're wearing the caps and the t-shirts: no reliance on trust. Evidence must be primary source. We're not Wikipedia, we're not looking for secondary source evidence, we're not looking to build on the trust shoulders of others. We are looking to build something that from first principles is defensible as being true. Evidence must also be open to all to evaluate. If it's only open to a few, it doesn't result in a particularly resilient system. It doesn't allow us to ensure that the system will perform ultimately as it's desired. And furthermore, it doesn't exactly build confidence in the system, even if those who create it believe that it should work exactly right. Unless it's built to ultimately be inspectable by all, it's unlikely that the broad population is going to have much confidence in it.
Again, this is a pretty basic principle, but I think it's worth pointing out. And we also have a principle of universality. Individuality is a universal concept. Either I am an individual—obviously I am an individual, but like as a digital avatar, either I am backing this digital avatar as an individual or I'm not. I might be an AI, I might be whatever else. It's not something where I am an individual in certain circumstances. If it's evaluated in the UK, I'm an individual; if it's evaluated in the US, I am not an individual. No, nonsense. Individuality, for it to be useful, it must be universal. Web3 is internet native, it is not UK native or US native, it is internet native. The internet doesn't care about where you are in the world. I mean, at the moment, anyway.
Okay, so let's move on. The nature, again, we have to find consensus on these three things, the three things I brought up before, the three things that we use to discover facts about the world when we are in a notionally perfect system. These three things, you can look upon them as a gateway from the imperfect into the perfect. Modern scientific method uses these three things to take the imperfect world, our measurements of the world, our inevitably imperfect, incomplete measurements of the world, and turn them into a perfect model. Now, inevitably, we work our way towards it. He's not going to deliver it on a plate and the first measurement that we take, but it is something with which we can ultimately hope to approach a quote-unquote perfect model of the world.
So we have to find consensus essentially on a new scientific method, but a scientific method for determining, for having a model of individuality. This is really, if we unpack everything, this is what we're trying to get at. So let's coin a term: a digital individuality mechanism. And we will coin this term as being a definition for these three things. Now, I gave you a rough definition for these three things in terms of a basic theory of knowledge that science more or less adheres to: collection, evaluation, judgment. And if we define these three things, we should be able to import individuality from the imperfect world, from our world, this lovely flesh and bone, metal and water world, into the very mathematical, very theoretical, very abstract digital system. The system itself probably isn't perfect, but notionally the internal model it provides should be perfect. If it's imperfect, we're doing our jobs wrong.
Okay, that's the ideology. Part two: theory. What does it need to be designed for? What are we really aiming for here? Well, we want it to be—I mean, the first thing that I'm going to mention, and I think privacy is something that is an element of Web3 that I would say is the earlier element. Web3, in some sense, is a combination of two different fields or at least two different communities. On the one hand, it's the one that most people here are most familiar with, which is the sort of Bitcoin, cryptocurrency, blockchain side of things. And on the other hand, it's more the OG privacy, communications privacy element. Now, both of them are about trying to provide guarantees across the board through engineering, mathematics for the interaction of people. One of them is just kind of seated very much in strictly communication, the passing of messages and the authentication of messages. The other one is in economic signaling, giving a strong economic signal again remotely with digital systems. And together, combined, they provide the elements for a platform to basically rebuild our digital systems from something where we have to trust people to something where we don't have to trust people so much, at least, and ideally not very much at all.
Individuality, for it to work, a digital individuality, for it to work, needs to be designed for privacy. Very basic. If it does its job but provides a huge database of, say, retina scans, that's not really doing the whole privacy thing. Justice. So firstly, there needs to be privacy when proving individuality. Whenever this evidence collection, however it manifests itself, it should not be that that evidence collection significantly impacts my privacy, or that I do not have confidence that it will not significantly impact my privacy. That's the first thing. And just saying 'trust me, it won't' isn't good enough.
Second thing is when that, however this individuality manifests itself within our digital systems, when it is used potentially in, let's say, two different contexts, it should not be possible to link that usage. This is crucial. Because as we know, metadata can be almost as compromising to privacy as the data itself if you have enough of it. And finally, if there is some sort of financial inflow, outflow, let's say, that comes about because of this individuality, then this should also not be inherently linkable, again for the purpose of metadata and ultimately for privacy. So the aliases that we would use in contexts must not be linkable, but nor must any payment accounts be linkable.
Secondly, it should be designed to scale. Individuality that scales to 10,000 people, that's not individuality. Now, Web3 technologies are, regardless of what people say, they are not ready to scale to 8 billion people all wanting to do Facebook over blockchain. It's nonsense. But over time, I think the technology will be developed to that point. And in the meantime, we still need to ensure that it can be developed to the point where a system of individuality could be delivering whatever this mechanism is, whatever this facility is, to everyone. And I mean, let's be pragmatic, it doesn't need to deliver it to everyone tomorrow, but it must ultimately have a path to get there, and it must be ready to deliver to a lot of people today.
It's rare for a human to have all the right answers off the bat. It happens occasionally, maybe even that human believes at the time that they know the right answers, but it's fool's gold. The ability to extend and adapt to some degree is important. Now, the ability to adapt can come at a cost of resilience, so it's a questionable trade-off, but the ability to extend need not. And therefore, it must be designed to be extensible. Whatever theoretical basis we might have for however we try to solve individuality, it's highly unlikely that the first solution is going to be the perfect solution.
So the first thing to do is to create common infrastructure that makes sense, common infrastructure that can stand reasonably the test of time, something that's minimal but fulfills its needs. The second thing to do is to ensure that there is at least one mechanism that defines these three points, how to gateway the imperfect world into the perfect world, and to ensure that other mechanisms can be added. And the final thing, possibly the most important of all, it has to be designed to be Web3, designed to work in an adversarial environment.
The world is becoming increasingly adversarial. Some elements of the world that were once thought to be not particularly adversarial are turning to become very adversarial, and there are plenty of new adversaries to add to the old adversaries as well. We cannot rely on all governments of the world, even governments of the world would agree we cannot rely on all governments of the world. And picking ones that we like and ones that we don't is really not a great way of proceeding. There can be no reliance on Web2 platforms, again it creates a trust issue. There can be no reliance on special hardware, how do we know it's been properly manufactured? And as I mentioned in the talk a month ago, it's not Web3 if there is a central keykeeper or a common confessional.
For those who don't know, a central keykeeper would be basically a Coinbase or whatever that keeps all your keys. Don't worry, if you want to spend money, I've got the keys, I'll just unlock it for you, hand it out to who you want, lock it back up again. Your keys are safe with me, sir. No, no. Common confessional essentially means give me all your information, don't worry, I'll only use the bit of it that I need to use and I'll keep it safe for you and I won't show it to anybody. Not at all, no, no. Well, I might do some analysis on it or if I get a legal thing then maybe I'll have to do that. And lots of small print. No, no. Trust. That's why we're here.
All right, building this system is a non-trivial endeavor. I can tell you, it is something that I've been involved with for one, one and a half, two, three years or something now. It's a non-trivial endeavor. Technology had to be built to get anywhere close to delivering a system that is in line with what I've just talked about. Let's look at the technology that had to be built. Polkadot-specific.
First thing I'll mention is a relay chain pub-sub. This is essentially the ability for parachains to very quickly, very cheaply publish data that is immediately available to all of parachains. It's a simple concept when you think about it, and it's something that you could in principle do over XCM, but it would become very costly and slow compared to how it can be if there's ground support for it in the relay chain. This is something we've built. Ring VRF. This is something that's used in Jam. It's a ZK-SNARK based primitive—there's a typo there—cryptographic privacy primitive, essentially allowing ring signatures with an additional VRF based on the Bandersnatch curve.
We have a fairly large change that I've worked on myself: transaction extension. Essentially, what this is doing is something quite clever. It's something that the Ethereum folks have been sort of talking about and working on for some time now in their own way. But what this allows us to do—at the moment, I hope people, most people, are at least a little bit familiar with this stuff. Transactions, when they make their way into a Substrate blockchain, they will typically have a signature. The signature on Ethereum, for example, is going to be like an secp256k1 ECDSA. On Polkadot, it's going to be an Ed25519 most of the time. Basically, it's a piece of cryptography that says I am a key holder and I've signed this thing. It identifies the key holder, and this makes its way down from Bitcoin, where you sign something generally to prove that you are the rightful owner of some funds.
It's all very well, but if we want to have something that natively supports privacy, particularly this kind of privacy for these purposes, we actually want to not have signatures, really. Because signatures immediately ruin our privacy. They attach us to a particular cryptographic key. We become not anonymous but pseudonymous. And as we know, being attached to a cryptographic key, if you're attached to enough of them, creates enough metadata that compromises one's privacy. Therefore, what we really want to be able to do is not have a signature associated with a transaction. Rather, we want to have a proof, a proof that does not identify us but merely guarantees that we have a property, in this case, a property of individuality. We belong to a set. We're not going to identify ourselves as a specific member of the set, but we will be able to prove that we are a member of the set.
Now, this actually hearkens back to something that Edward Snowden brought up five years ago, might have been six, I think it was five, at this very, in this very room over video link, where he mentioned that we are getting to the point now where it's not enough to have money to spend money. You also have to belong to the right group of individuals, or rather, not belong to the wrong group of individuals.
This, I mean, it begins with a thin end of the wedge, right? Okay, we don't want an undesirable politician here. I'm not going to get into modern global politics, that's a losing game. As long as you're not that person, you can spend money. Yeah, this seems kind of reasonable. Okay, well, we're going to extend it to this group of people. It kind of seems kind of reasonable. Okay, well, we're going to extend it to everyone who's had any kind of link to the... and it's like, eventually, it becomes a problem. It becomes a problem partly because it's easy to add people to the set, and so it becomes done for the wrong reasons or on a whim or with bad evidence.
And we saw this when, for example, it became hard to donate money to the legal fund of a certain whistleblower in our era. Basically because Visa, Mastercard, and PayPal were asked very nicely to block this person. And this is becoming the standard, unfortunately. It's also problematic because in order to enforce this, it's not enough—we don't, it's not like the governments of the world enforce this through cunning zero-knowledge proofs that guarantee our privacy, just make sure that we're not in this set. No, no, no. They enforce this by saying, well, whenever you check into an Airbnb or do an OTC Bitcoin transaction, you need to provide your passport and they need to write it down and then they need to submit all of these passports on a digital system so the government knows exactly who is staying where and who is buying money.
This is not exactly the society I want to be living in, to put it mildly. Let's move on. More technology that we built. One of the things that's been in the works for a while is something called the bulletin chain. Now, essentially, this is a very easy, very cheap to run piece of decentralized secure data storage that is heavily integrated into XCM and by virtue Polkadot. Essentially, it allows Polkadot to say this account or this data, the pre-image to this hash, is allowed to be stored for free for two weeks.
Yeah, this is really handy. And bulletin chain is also improved a little bit further by the fact that the data can be both uploaded easily using web technologies and downloaded easily using web technologies. Second thing is the pallet that runs all of this runs out of core infrastructure for registering and utilizing the concept of individuality, in particular registration. And all of this key management, building ring signatures, yada yada yada. This is, you know, not entirely trivial, and that's what this pallet does. And finally, there is a pallet that we call sort of the vouchers pallet, and this essentially allows treasury payments that do not immediately, that do not link, or at least not fundamentally linkable, to the reason for treasury payments.
Yeah, now this is crucial to unlink any payments that the treasury wants to provide on the basis of individuality, or something stemming from individuality, from the individual themselves. Basis of privacy. Now, I mentioned this a little bit last month, but in terms of API, the point of being able to do this is to be able to switch out, in whatever parts of your codebase that you want to switch it out, that makes sense to switch it out, a line. Now, you'll find, if you're a pallet developer on Substrate, on the parachains or whatever, you will find in many places a line that says let account or who, or whoever it is, equal ensure signed and then the origin in brackets. And what it's doing is it's saying, whatever, wherever this transaction seems to be coming from, I want to make sure that it has a signature attached and I want to know who signed it, and it gives you an account back.
Yeah, and what we want to do is say, actually, I don't care if it's signed. I don't want it to be signed. I'm not looking for a signature. I don't care who it is. I only care that A, it's a person, and B, I want an alias given my context. I'm not trying to do it for anyone else's context. I'm not trying to discover who they are in some other context. I care about knowing that alias in my context. Why do I want an alias? Well, because I don't want the same person appearing as two different things in my context to me. That's, you know, if that were allowed, it wouldn't be sybil-resistant, and the whole point of doing this would be gone. Same API. There's some cunning bridging to be done, but that's about all. And this is something that we want to see also rolled out into other ecosystems. Now, we have two high-quality bridges now between Polkadot and Ethereum, and it makes sense to start ensuring that these ecosystems can become as symbiotic as possible and as helpful to each other as possible. So further integration, well, you know, we have...
We want to be able to have multiple digital individuality mechanisms, and part of this is essentially exposing an API for when a digital individuality mechanism says, 'Well, look, I've discovered an individual. I am really confident that all of the other individuals that I've brought to you, this isn't one of them. It's a new one, and for any others, I'll make sure that it's not one again.' That's what we have. Citizenship really recognizes a person. Simple API. We of course identify them by a key, but it's a band snatch key, right? It's a key that allows us to jack into this ring VRF and ensure that they are not individually identifiable when these proofs arrive when they're using this key.
So as I mentioned, there's this new cryptography that we're using. It's already in use in JAM and Saffas, and this is something for wallets and in chains. It's relatively easy to integrate. Wallets will, if they want to integrate this, it's still relatively... Hardware wallets will obviously have a bit more of an issue, but it's unlikely to be that necessary for hardware wallets. It's not like this key is going to be storing significant financial sort of assets.
