Hashing It Out
Hashing It Out

Episode 114 · 4 weeks ago

Hashing It Out Personals: Ari Juels

ABOUT THIS EPISODE

Today we’ll be talking with Ari Juels. Juels is Co-Director of the Initiative for CryptoCurrencies and Contracts (IC3). He is also Chief Scientist at Chainlink Labs and a Professor in the Jacobs Technion-Cornell Institute at Cornell Tech. We'll be talking about his work and what motivates him.

You can watch this interview on YouTube .

M H. Welcome back to hashing it out. This is a personal interview. Today we have Ari jewels. Um, we can just do the normal thing and give us a quick instruction as to who you are, work on and maybe how you got there. I'm a faculty member at Cornell Tech in New York City, on Roosevelt Island. I'm also a CO director of I C three, or the initiative for cryptocurrencies and contracts, and finally, I'm chief scientist of chain link labs, which is a company that builds oracle technologies for our block chains. We were just just chatting a bit before Um we started this and we were discussing you're you're mentioning the like massive over requests to come to SPC this year, which is why I would consider a Premier Technical blockchain conference where most, I guess, of the Um academic work suits tries to get there and and talk published discuss recent findings. Can you can you talk a little bit about this conference and maybe the history of it leading up to where it is today? I remember correctly, it was never this popular, or hasn't been this popular in previous years. Well, it's a relatively new conference. I believe the first edition was in run by benedict boots and den Benet. It was initially called the Stanford Blockchain Conference and Twenty Nineteen. It certainly a pretty heavy attendance, but I think we saw the first stirrings of Covid at that time. So many attendees, or registrants rather, ultimately didn't attend. We've changed the name to the science of blockchains conference and Stanford is now partnering with I C Three To run the conference next year. The conference will be taking place in New York City, most likely in the fall. This year we had something like four thousand registrants for a venue that can only accommodate one thousand. So unfortunately we had to turn away many would be eager participants in the event. But it's a it's a it's a wonderful, lively event with a wide range of talks, many of them quite technical, some of them are not so technical, but all of them, in my experience, quite interesting and engaging. I've been in, I guess, the quote Unquote Change Space for a number of years, maybe conditionally got started doing, like I guess, legitimate Um studying and understanding around two thousand and eleven Um, and I remember in the early days the academic setting, if you will, was was scant. Getting people to understand with, I guess, with Rigor Um, how blockchains are created, why they're useful, with the limitations of them, are the different components that build them and their individual limitations, was very difficult and over the years I've seen that dramatically improved, and I think that has to do with they're starting to become some level of acceptance within university curricula. Can you talk about, like, the difficulty of trying to introduce something like three blockchains into standard computer science curricula over the years, because they in my opinion, it's always difficult because the technology is moving too fast. You don't have any regularly cadenced coursework. It makes sense it is a difficult topic to teach precisely as you say, because it's so fast moving. Any textbook is going to become obsolete within a year and it probably takes a year to get a textbook published. The relationship between the blockchain community and the academic one is quite interesting actually and extends back some decades. Really early in or in the nineteen eighties, people where the research community was exploring concept called the Cash David Chum for instance, had some landmark ideas, for instance, Um, what are called blind signatures, which were instrumental in constructing privacy preserving digital cash schemes, and this was a subject of academic study for quite...

...some time, through the nineties, for instance, Um. But ultimately that venture failed. E Cash, which was very much like cryptocurrency but involved a model of centralized issuance, just ultimately didn't go anywhere for a number of reasons. One was the failure of one company that played a pivotal role in the area digit cash. Um. Another was, I think, that the technology was a little bit of ahead of its time, and so on and so forth. Okay, roll forward to two thousand and eight and this wonderful, insightful, brilliant construction just drops out of the sky. Of course, we still have no idea who sets Nakamoto is, but Setoshi Nakamoto, wherever she is, he is or they are, proposed a way of constructing digital cash. I'll call it using the termam it's broadest sense. Um In using means that the academic community had had just not conceived of this combination of a decentralized system and mechanism design and so on and so forth to build a digital currency system. Um Really, I think, took the academic community by surprise. But at that point I think interest in the CAS should fizzle out a bit, and it wasn't clear that Bitcoin was going to take off. So indeed, it took some time for academics to recognize its significance, both both its intellectual significance and the real world impact it would ultimately have. But it's worth pointing out that even some of the poor components of Bitcoin had their roots in the academic community, of formal research community, if you will. The Hash functions used in Bitcoin, for instance, shots to fifty six, right, m d one sixty, just to give you an example, not to mention the digital signature schemes I had, had their origins in academic research. So in a sense academic research played an important role in the development of cryptocurrency since the beginning. That said, interest from academics and Bitcoin was pretty muted until, I don't know, twenty eleven, twelve, when I think, it began to everal world impact and people began to see that it was addressing the problem of consensus in a profoundly new way. And now I would say that the academic community is appreciating and having an influence on the blockchain community in a way it's almost unprecedented for academic science. I worked in applied cryptography as a researcher for some time and the work I did over the course of many years saw a very little uptake until I went into academia and started looking at blockchain technology. To answer that's a long way to answer to your question with teaching this stuff. Um, I do teach a course, couple of courses actually, on blockchain technology and indeed it's hard to create a stable core curriculum. But on the other hand, the fact that things move so quickly keeps the material fresh and makes these courses very lively and keep students engaged in a way that I haven't seen in teaching other courses. MHM. I come from a background of robotics. Specifically, I I went to school for Electrical Engineering and at a young age I, you know, was fascinated with computers and I wanted to be able to build things, you know, Legos was kind of like my um my inspiration for wanting to put these these computers together and make them do things. Um, it sounds like from from your history. Um, you were into cryptography before, you know, any of these digital cash mediums existed. How did you get into that sort of work and study? Yeah, great, great question. Actually, it was digital cash they got me interested in applied cryptography back when I was in Grad School. Fellow Grad students of my name, of Pete Demo, gave me a manuscript, I think it was, and I don't even remember whether it was a published paper, by phone name Stephen Brown's, who had devised the cash system with a broad range of functionalities, and I was just enthralled by this thing because I had not encountered cryptography before and I didn't realize how the into which math could...

...be used not just to model the world but two construct new protocols, that it had this kind of constructive potential, which it does in cryptography. Numbers matter in an essential way, in a fundamental way. Um, that that's that looks very different from the way, for instance, they're used in and let's let's let's say physics, more as a way to model things or to guide the construction of new systems. In the case of Cryptography, the numbers are the system, so that this this was. This was many years ago. I was in Grad school exploring a completely different area of computer science, something more akin to machine learning and cryptography. But when I graduated I wanted to find a job and apply cryptography and found the position at a company called R S, a company creating applied cryptography, cryptographic tools. So, as I say, actually was it was digital cash that got me into applied cryptography to begin with. That's wild said. I come from a background of computational physics and I think it's I was first introduced to the concept of Cryptography, maybe through books are reading, but realistically through information theory, treatment of thermodynamics, which taught me about, you know, encryption, decoding, and you look an interview through information there you get this concept of Um encoding and so on and so forth that you overlapped quite well with a lot of cryptography. But I hadn't thought about the use of it, like you just mentioned, and that Um it is much more involved and useful for a given scheme than math being used in other disciplines like, you know, computational physics or chemistry, and that it's a model to help describe the universe, whereas Um how you implement a given type of cryptographic protocol changes the way the humans are interacting with each other and what they rely on, what they trust, and what they have to do and what they have they don't have to do, and I think that concept is relatively new to people. Um, we built the Internet the way it is today without people having to worry too much, but they can do more stuff through some of these protocols. Are Saving one of them in terms of them being able to put their credit card on the Internet, and how we make new protocols, specifically kind of these blockchain networks, is allowing people to transact money or other digital scarcities, and I and I think this concept is relatively new to people and that we can make computer systems built on math that allow humans to do new things, specifically digitally. Uh. Is that resonate with you at all in terms of like just like new concept of how we each math and how it changes human to human interaction? Yeah, it's a strange, encounterintuitive thing, this building of systems where the building blocks are hard mathematical problems, and the fact that these systems, the way in which these systems enable human beings to conduct commerce in a trustworthy way, all all of these are are as, I said, this kind of novel and counterintuitive concepts. I think yeah, so, so I think your your view does resonate with mine when you when you're talking about how your interest in cryptography sparked from digital cash. What was it about digital cash? Was it? Was it partly beyond the mathematics? Was it partly the economics, like the the look the potential for it to be a lucrative field in the future? At the time I was more interested, I think, in the applied cryptography, in the paper I was reading on the cash, then the cash itself. The fact that enabled this application was somewhat interesting, but, as I said, was the fact the idea of building a system where the building blocks are hard mathematical problems that was so intriguing to me. And digital cash, the ability to transfer value on blockchains, is certainly interesting, but it's not what interests me at least most when I think about blockchain systems. I'm more interested in smart contracts and some of the more advanced functionalities that blockchains...

...enable than I am simply in the ability of human beings to send money to one another instantaneously and quasi privately, although that is obviously important and impactful. What is these other applications like? How do you see these networks expanding beyond simple value transfer? That so simple value transfer? Yeah, it's it's really hard to say and that I got involved fairly early on in my foray into blockchain land. I got interested in Oracle Systems because I recognized that if we're going to do something really interesting with blockchains, with smart contracts in particular, at least interesting in my view, those smart contracts are going to have to consume data from off chain systems. And at the time I started thinking about this kind of thing, back in Twenty fifteen, say, we didn't really have great oracle system designs. As I began to explore the design of Oracle Systems, I started to encounter some interesting use cases. Um Defy certainly, is certainly one of those, and Um. Another that has been particularly of interest to me recently is U N F T S Um, which of course can't be constructed independently of the use of oracles Um, but which I think are going to be greatly enriched through the use of of oracles Um. Future use cases are are kind of hard to anticipate. People talk about things like insurance and automated insurance and so on and so forth. Um, certainly think that applications of that kind are going to be interesting but um, it's really hard to know where the technology is going to go. Two years ago I would not have guessed that images of of apes would be one of the premier use cases for blockchains. And on the face of it it's a pretty frivolous one. But I think actually it's going to have a n F T is are gonna have a profound cultural impact. So, in a word, I would say at the moment at least, u n F T S are one of the applications I'm most enthusiastic about it and and defies another. I just think it's it changes the way we conceptualized financial instruments. So fundamentally I think it's going to have profound and lasting impact on the whole financial sector. Do you like? Because we were maybe unable to see, say, we'll say four years ago, we weren't able to see the kind of proliferation of n F T S and how it got everyone excited. So we were worried about I C O S and regular, you know, just making coins. I'd say at that time. Um, if you had to look back and look at Bitcoin to today and the massive swath of platforms and applications leading up to maybe doals and N F T S, is forward thinking applications of these things. Like what do you what are the what are blockchain is useful for? Did you have simple value transfer, digital scarcity, right, like fungible digital scarcity with Bitcoin, and then we generalize that with ethereum. Two be somewhat arbitrary. Scarcity um not necessarily attached to and just just value, but metadata. N F T S are just basically a token of of metadata and ownership and then some value. Like what is what is the general purpose for blockchains? Why? Like, why? Why are they useful and what context? Well, if we look at Defi, as I said, I think they transform the way that we conceptualize financial instruments. They now become computer programs instead of legal agreements, which is essentially what they are in the real world. Now, of course, legal agreements. There has to be an interface between legal agreements or legal convention and blockchain assets, but this transformation to computer programs, I think, is going to have an enormous impact. Let me let me give you one example. My favorite example is flash loans. There is no analog that I'm aware of too. Flash loans this blockchain phenomenon in the contional financial industry, and it is the technical properties of blockchains that allow these things to exist. So, uh, just for listeners who may not be...

...familiar with flash loans, flash loan is essentially the ability to borrow an essentially arbitrary amounts of money with no collateral and without identifying yourself. The catch is that you have to use borrow the money and use it within the same transaction. If you borrow the money and you fail to pay it back, the transaction is rewound, it reverts and it's though as though you never borrow the money to begin with. But try walking into a bank today and saying I want to borrow twenty million dollars. It's just to the day, but I don't want to provide an collateral and I don't want to identify myself. Right. You would be quickly, I think, ushered out of the bank. Um, if somebody pressing a red button under a desk somewhere. So that that's an example of one of the capabilities that this reframing of financial instruments as software enables. And I think there are a lot more, and this is actually one of the topics that my group is exploring it Cornell Tech. We're thinking about what other novel financial instruments you can construct using blockchain technologies. That's one example. Uh as for N F T S um you know, the moment they're basically just tokenized jpeg files, but I think they can be a lot more, and let let me give you some let me describe some of the research my group is doing, and I think that will help illustrate some of the ways that the n f t landscape can be enriched. One of the problems the N F T communities and capturing today happens when n f t drops take place. As you know, N F T S are often sold in big batches like and what often happens when an n F T is popular is that at the time it's dropped, the N F T s all get snapped up by box and the BOT masters then turn around resell these things, the essentially scalped N F T s. This happens also for sneakers, you know, all kinds of popular items that are sold in this way through drops. How do we remedy this problem. I think that the key people have tried to do this by restricting a number of N F t s you can buy per address and so and so forth, but those are federal approaches. So I think the key is going to be decentralized identity. My group did a little experiment a couple of months ago at East Denver. We ran not an N F t drop but a raffle which looked very much like an an n f t drop, in which we enforced the policy that a single user could receive only a single raffle ticket, equivalent to a policy and an n f t drop where you allow only one N F T P purchase per user. And the way we did this was by having users create essentially decentralized identity credentials using, for instance, their social security numbers in a privacy preserving way. Is Important, of course. So now we're talking not just about tokenized ape images but about a an ecosystem of collectibles and art interfacing with a decentralized identity ecosystem, and this can go much further. Our ultimate goal or vision is that artists should be able to create through a kind of policy engine, smart contracts that will enforce a rich range of policies on WHO's able to buy their N F T S, how they get resold and so on and so forth. Some of the artists we've spoken with our interest in offering discounts to fellow artists. How did they do that? There's no way to do that today, at least. We out just relying on the honor system, which is going to be a fragile and blockchain in the blockchain ecosystem, of course, but if we have a decentralized identity ecosystem, then that type of thing becomes possible. MHM, it looks it's it just sounds like you're really researching a lot of relationships that can be had with the agents who exchange these N F T S. I know personally I try to use n F t s beyond Um just simple, you know, tokenization of JPEG's in a way where it was um allowing users to kind of level up in some sort of tier system within a a Dow Social Institution where you could accrue some sort of reputation within the system by participating with non financial incentives, but these tokens would be given based on your ability to help. You know, each other out and to communicate pause satively within a community Um. So, like I was very...

...interested in building Um and leveraging N F T S in that sort of direction. So I can definitely see, Um, where your interest comes from. When you when you're talking about N F T s having a use beyond, you know, just treating pictures. Yeah, you're you're basically talking about decentralized identity, credentials, reputation in the sense is credential, and that reputation can be tokenized in the form of an N F t. You can think of pollops, for instance, are not quite a tokenization of reputation, but a tokenization of some form of interaction in the community, for instance attendance at conference. And indeed, N F T S can represent all kinds of things. Plus they can represent fine art, they can represent a range of collectibles. Ultimately, when we all migrate to the METAVERSE, whatever the metaverse is, I don't think any of us knows, are possessions once digitized may well take the form of n F T S, and that means they're going to be of really deep importance in our day to day interactions. I have an interesting side question here. Like you, have a good concept of how things currently work and what or reasonably good research areas on how these things can be used in the future. Why stay in academia? Why not build a company and and flesh out these concepts outside of I would say reasonably constrained environment for I don't want to say innovation, but I would think, I would argue academia tends to move at a slower pace than than industry. What keeps you there? What? What? What motivates you to to to stay in this environment and not break out and kind of work on your own? That's a that's a really good question. I spent seventeen years in industry and the reason I moved to academia was that I felt I could have a greater impact on industry practice from academia than I could in industry, and that's actually turned out to be the case in a way. So academia is perhaps slower moving in terms of its core curriculum, but in terms of the research at academics do, it's much faster moving. Faster moving in the sense that we're exploring ideas on a time horizon well in advanced often of that in industry and startups, for instance, are just trying to get a product out in the next three months and not thinking necessarily about what that product is going to look like in five years. Often there is a gap between the research academics are doing and both current and conceivable industry need. I don't think that's true in the blockchain community and that's one of the reasons that I'm doing research in this area. Academic ideas are being translated into blockchain systems at an an incredibly rapid pace, sometimes alarming, sometimes, yeah, indeed, sometimes alarming, because often the technologies that are getting deployed are not terribly mature. But I would say there's a wonderful synergy between the industry blockchain community and the academic blockchain community and in fact one of the reasons that we founded I see three was to facilitate the conversation between the two communities, which we knew even at the time that we launched I see three back in twenty six team, which we knew was going to be um a vigorous one. So Um Academia, I think, is a great place to explore blockchain technologies in a freewheeling way and to have a substantial impact on industry practice. I have also, as I mentioned, one of my roles is as chief scientists to chain link and some of the ideas that my group has developed on the academic side are being taken up by chain link, some Oracle Technologies and technologies for enforcing what's called fair transaction ordering, trying to level the playing field in Defi as it work. I would I said, I definitely agree with that from my experience of of academic fields compared to industry. Think it's it's it's it's exciting to see the level of integration between...

...the two and how fast some of these things get picked up, implemented and then Um kind of helped prove whether or not something works, or at least tests, like test the boundaries in a real environment and whether or not something works. Because when you add almost the cases of an application within blockchain is monetary has some monetary and centers associated with it, when you add that, you end up with a lot more motivation to break things than then most experiments, I'd say. So you end up understanding whether or not something is works the way he thinks it does, because people want to go through anything that can possibly do to break it in order to capture money from it or gain it in some way, and so like fat aspect of motivated attack and tight integration between what I would say vigorous development of technology is something that they think is a wonderful combination and I'm aded to see it kind of grow over time. Do you see the pitfalls or things to look out for in in in the future of that developing tighter and tighter? Yeah, that's an interesting question. Um, I would say that money can have a corrosive influence and I think can distort people's interest in the new technology in ways that are that are not always healthy. So I have seen students who are interested in blockchain technology because they want to make a quick book, because it's so intellectually engaging and because it's interesting. Is One of your students was Phil Dian, who kind of created the concept of MTV in the first place, which is now motivated a good portion of this, I guess, subculture of people looking to take advantage of transaction ordering. Well, yes, that's an interesting history there. Um, so, as you suggesting, you know, my group published a paper, Phil was the lead author, on what came to be known as MTV. That's the term we coined in the paper back in and since then, the PhD students in my group have taken, I think, the divergent paths towards addressing the problem of MTV. Um, I think we all in the end have the same goal, which is tried to try to mitigate the negative impacts of MTV, but we're trying to get there in different ways. So Phil in the CO founded flashbots with the idea that MTV is unavoidable and the best we can do is to democratize it. And I'm somewhat simplifying his his point of view here. Um, that's not my prefer it approach to the to the problem. I would prefer to see US start with an approach that tries to reduce the negative consequences of MTV from the outset. So I have other students looking to build what we call fair transaction ordering systems. My student, uh the Kel car, for instance, has been spearheading work in this area, helped by C Sean Long we have built a system called them, this successor to Aquitas original system that looks too in a nutshell, to order transactions essentially on the first come, first serve basis. And it's my belief. This hypothesis still needs to be proven, but it's my belief that this will lead to fairer defy systems in the sense of reduced seeing the harmful impacts of MTV, reducing MTV opportunities. Well, we'll have to see whether that indeed comes to pass, but the system Famnis, for instance, is being taken up in industry and I hope we'll we'll roll out soon. Do you feel like there's a there's an ethical dilemma for working in the space and seeing some of the theoretical Um works become I guess, like what was it Mount Olympus or Olympus Dow? Um, like, I guess some of these defi projects that end up kind of causing a lot of people to lose their money. Um. Yeah, well, this issue is ethical issue. Isn't peculiar to blockchain technologies. All technologies a double edged sword and Um, it's...

...been very encouraging to see the range of interesting and positive use cases for smart contracts, discouraging to see some of the abuses. For instance, my Um Group, along with Saryl, Sarah Michael John, Um very well known researcher Um in particular, who's done seminal work exploring the privacy limitations of Bitcoin. Um We've explored a smart contract called forsage. forsage, I don't know if you've come across it at all. At one point it was the number three contract in ethereum by gas consumption. It was a pyramid scheme and it was pyramid schemes have been around for forever right Um. But it was facilitated by smart contracts in the sense that the very trust properties of smart contracts could be used to market this financial vehicle in a misleading way, and in particular those Um promoting forsage suggested that, because smart contracts are autonomous and have all of these trust minimizing and so on and so forth, that forsage couldn't possibly be a scam. But it was essentially a classical pyramid scheme where money flowed to the top, nearly to the in particular to the creators and promoters. And, as you may have seen, the the SEC now is is prosecuting both the creators and some of the more prominent promoters. So you know, the fact that smart contracts can be used for this purpose is somewhat discouraging. But but I think inevitable for any technology. So we as technologists certainly have to reflect on the uses of technology, how we want to promote it, how we want to steer it in an ethical direction. But there's no such thing as a technology that can be used purely for for good. That's true blockchain technology. That proves proves true so many others. Certainly in the case of MTV um there's there are a host of really thorny ethical questions which we have we debate vigorously within my group and I don't think we've come to any conclusions about the best way to approach the MTV problem. I would imagine that also stands for privacy preserving technologies as well, particularly within blockchain or not. And then one of the recent larger controversies, slash influencing things going on is the sanctioning of Tornado cash uh by the United States and then people like Matt Green Republishing Them for teaching purposes or, you know, code speech scenarios like. How does like, how do you see that playing out? Because with the current ongoing innovations of privacy preserving technologies becoming more robust, like, you know, via zero knowledge applications or whatever, whatever you can come up with. Mostly, I says your knowledge applications right now, like the ability of people to do surveillance on a network is going to get more and more and more detrimented. Um and policy people are not going to like that, or at least you know, people who try to enforce criminal justice, solicit finance activity whatever, are gonna like that. But I don't like I don't see. I see this clash continuing and growing over the course of the next five years. How does how do we, as people who make this stuff, hopefully ethically, like push forward in an environment like that? Yeah, this is this is certainly a really challenging question. Paradoxically, privacy preserving technologies can erode privacy. BITCOIN is a case in point where, ostensibly it's a privacy preserving technology, turns out that it doesn't provide real anonymity. It only anonymous bitcoin network. And Yeah, exactly how the the I guess, consequences of people acting as though it was exactly right. Private privacy enhancing technologies can load people into a false sense of security and in Cybersecurity in general the advantage lies with...

...the attackers. So I'm not convinced that rolling out even stronger privacy. Preserving Technologies is ultimately going to result in stronger privacy, for good or bad for for the users of these of these systems. But in any case, what I do think is important is striking a good balance between the need for law enforcement and regulators to have insight into what's going on in these financial systems and the need the right for users of these systems to enjoy some degree of privacy. I think we can try to strike this type of balance by building in accountability mechanisms. Doing that as hard. An accountability mechanism often takes the form of a back door, and back doors are slippery slip in all kinds of ways in terms of the technologies that implement them. Wants to create a backdoor on the risk that's going to get compromised and politically. If you make the back door available to one UH government, then other governments are going to clamor for exactly the same access, and some governments may be institutions in general, may be more well intentioned than than others. So I think that these questions in general are are pretty thorny. But I do think that by creating, let's say, Um more levers, more sophisticated ways to access secrets, we may be able to help ensure that regulators don't take too heavy handed an approach to blockchain technologies. I think regulation is very important. Consumers need to be protected, but they need to be protected in myriad ways, and one of those is having their their privacy protected. In the case of Tornado Cash Um, the approach was, perhaps, out of technical necessity, a pretty heavy handed one that caught up, I think, Um, a number of perfectly honest, well intentioned users. Um, if we build systems, as has suggested, with more levers, and it's not entirely clear what form these, these levers and nobs should have, but but with more Um flexible technical capabilities, then we may enable regulators to take a more nuanced approach to solving the real problems in the blockchain landscape, cryptocurrency landscape. When you when you mentioned that you one of your research groups was looking into applications of decentralized identity. Does that? Does centralized identity um include, I guess, some level of privacy guarantees so that your identity is rolling or your identity is hard to map to, you know, your real identity. Um. I've had experience with bright I D in the past and some you know, Um, sort of like decentralized identity products. But could you talk more about, like, I guess, what what level of privacy guarantees are associated with, how you envision good to centralized identity means? This is this isn't a topic in which the academic community, or at least that the scholarly community, let me say, it's it's done a huge amount of research and there are a whole range of credential systems, as as they're often called, that Um provide any of a number of different privacy mechanisms. So you've got anonymous credentials, you've got credentials with conditional nentymity can be revoked by a committee, for instance. Um There are um credentials where your nentymity can be revoked if you abuse the credential by using it, by double spending, for instance, using multiple times in an appropriate way. So we have lots of technical tools here. The problem that my group has been focusing on is relates to privacy but is slightly different, and that's the problem of bootstrapping that decentralized identity ecosystem right. So there, there, there is a large spot of the industry interested in decentralized identity. Even Um, let's say, well established technology companies are interest in decentralized identity...

...because they don't want to incur the liability of storing highly sensitive data. They would rather see that that data in the hands of users. We have this idea that big tech wants to suck up all everyone's data as much as possible, and this is true to a certain extent. Data mining is obviously a profitable business, but there's liability associated with holding sensitive data and there are many firms that would rather not have to custody that that data. So there's broad interest in decentralized identity, but there's a fundamental problem here, the chicken and the egg problem, namely that, Um, we can't create a decentralized identity system until there are issuers, as there are authoritative entities around that can create decentralized identity credentials, right until the D MV is willing to issue you a driver's license on a on a block chick. or Or is it another type of decentralized identity credential or reputation tokens, which you were suggesting earlier? Jesse are issued by some some authority. So, Um, we won't have a decentralized identity system until these authorities, these issuers, exist. But nobody wants to create an issuer or run an issue and service until people, until there's a use case or decentralized identity. Right. So, as I that we had this chicken and the egg problem, and this is the problem that my research group has been taking aim at Um. UH, two students of mine, particular, DPAC Marm and Fan Jang Um. We're UH spearheading the development of a system called Deco, and earlier we had a system called town crier. These oracle technologies uh, that we feel can be used to address this problem, and the way that they get around this log jam is by creating decentralized identity credentials based on existing web data, with no modification to existing web servers. So so here's the idea. As I said, the D M V is probably not going to become a decentralized identity issuer anytime soon in an explicit way. But suppose that we can, in a trustworthy way, export data from D M V websites onto a blockchain. Well then we can create the centralized identity credentials without the D M v having to provide explicit support about they're necessarily even knowing that this is happening, and using these oracle systems that I've mentioned, that that is possible actually and we can do this in a privacy preserving way. So, rather than just exporting all the data on your driver's license in its entirety, and that may include your birth date, which most definitely don't want to expose on a on a blockchain, what we can do is export bits of data piecemeal or functions of this private or sensitive data. So, for instance, you can create a decentralized identity credential on the blockchain saying proving that you're a resident of California according to the California d M V, without also exporting your your birthday. So, as I said, it's this chicken and the egg problem that we've we've been taking in that with these oracle technologies and along with Jaslene Malvi Um, these two students I mentioned, have Um lead the development of a system called candid, which uses oracle technologies to create decentralized identity credentials of this way, and also in a privacy preserving way to de duplicate identity credit atols to enable you to create a unique credential. So can you can accomplish things like enforcing a one N F T per person policy in an air drop. Interesting, slight, slight change of of of questioning here. Um, I don't know, maybe not. How do you teach people about this? I'd imagine one way in which you've experimented with and others have, is writing novels around how the stuff works and what a realistic scenario or world looks like had these things been implemented correctly. Um. Some write more dystopian pieces, like Corey Doctoro. Others don't. Uh. Like you wrote one book. Can you talk about that and why you chose to write it and if that's like a ongoing process for you, of of of writing novels to help build worlds on how this technology uh, can...

...change. Yeah, I wrote a novel entitled to Practice Number of years ago now, uh, and that the novel relates to our early discussion, earlier discussion about Um Mathematics as a building block for computer systems and through through cryptography in particular, and the premise of the book was that a Pythagorian cult. Pathagoras, of course, was an ancient Greek philosopher and mystic who believed that number was numbers, as it was the building block for the for the universe, Um, not, not just for computing systems, because, of course there were no computing systems at the time. So the premise was that pathagorine cult had figured out how to factor the product of large primes and to break the UH mathematical problem that underpins the R S, a algorithm which is a fundamental cryptographic primitive used to secure protocols on on the Internet, more prevalent at the time that the not norvel is published than than now. But there are alternative mathematical basis for the cryptographic protocols we used today that are vaguely similar. So that, so, that's what the that was the prepose of the book at the time and it was an exploration of how cryptography is changing the trust equations according to which we conduct commerce and our lives more generally. Um, as it happens, I've been working on a novel about blockchain technology. Um, won't give any more details quite yet, but I have just finished up the manuscript and I'm shopping it around now. Um, and Um, I think, Um, I've learned a lot about writing since I wrote the first book that this one I'm much happier with than I was with detractice. And this book is based in large part on the research that my group has has been doing. If you're feeling FROGGY, I'd love to read it. UH, happy to give you a copy. Yeah, so I we we interviewed Carl schrader a while ago who wrote stealing worlds, which was, in my opinion, one of the better depictions of a universe Um, where, I'd say, alternative economies fueled by technology and are allowed, for like, place a people of place for people to opt out, and gave a realistic scenario of how this stuff kind of could change the world. Um, I'd be curious to see how other people building universes around this techno, aology also kind of view that and whether or not it's it's what implications exist and how they view the pitfalls and changes in human relationships and so like. If you haven't read that, I would definitely encourage it and also I'd love to read what you're writing. Yeah, be happy to happy to shoot do one. Yeah, Corey. I've read a lot of corey doctor's work too, if we've interviewed him a few times, and it's it's definitely, uh, much more in your face of the dystopian point of the things that we're building and what negative implications they may have. Is your is your worldview more optimistic? Um, yeah, I would. I would say that. Yeah, yeah, law, and pause there. I wouldn't characterize my book as Dystopian, although it does touch on some of the ways that blockchain systems can be abused. Uh, I would consider it to be a pro blockchain work in the end, and the protagonist is certainly pro blockchain, far, far more than I am, in fact. How do you in your experience of writing novels? Has that like? That's a difficult process. It's a very time consuming process. Um, it does it? Have you found that experience to change the way you think about technologies and and like as the process of exploring and flushing out narratives and implications of those narratives and building the world? Has that changed like research directions or your understanding or thoughts around a given technology and its amplications? Well, in some sense I'm in my main profession as a faculty member Cornell Tech in the business of writing science fiction, just that I happened to...

...publish it in solidly Venus most of the time, um. So I consider writing novels to be an extension of, or compatible with, the work I do on a day to day basis. As I mentioned, the novel reflects some of the researchers going on in my group and I in some ways it's it has um inspired me to dig more deeply into some of the topics we've been exploring because in writing the novel I realized how important they can be in this Um fanciful future that I've crafted in the novel that may become reality in five to ten years. So there's certainly a conversation going on between my literary endeavor here and my work as a as a researcher. Um. Writing is hard. It's so it is. It's hard for everyone, but for me at least, it often becomes a flow activity, if you're familiar with that concept, the type of activity you can plunge into in a way such as you lose track of time, and for some that's the that is the an optimal psychological state. That is the definition in a roundabout way of happiness, I suppose. So Um it's a it's a it can be an immensely frustrating process, but it can also be a really fulfilling one, for me at least. Think that's awesome. Is there you to wrap up? Is there any questions you would have liked us to ask that we didn't? I think this was a nice, wide ranging conversation, so I don't have any other questions or information to impart Um. That's top of mind at the moment. Thank you very much for asking such interesting and incisive questions. I've really enjoyed this conversation. Thanks for giving good answers. Yeah, thank you. Thank you.

In-Stream Audio Search

NEW

Search across all episodes within this podcast

Episodes (118)