The Infrastructure Beneath the Next Economy

Join us as we speak with Douglas Heintzman, Chief Catalyst at the Blockchain Research Institute and CEO of Syncura, about how digital trust infrastructure is being developed for real-world use. This episode explores how blockchain supports trusted data, how identity and credentials can be verified in real time, and how these systems interact to reduce friction in transactions. The conversation also covers governance, interoperability, and how infrastructure-level trust can support more efficient systems and broader adoption across industries.

Disclaimer: These podcasts are produced by the Cardano Foundation. The views expressed in these podcasts are solely those of the participants and do not represent the opinions of the Cardano Foundation. These podcasts are for informational purposes only, and are not intended to provide legal, tax, financial, or investment advice. Listeners should consult their own advisors before making these types of decisions. The Cardano Foundation does not guarantee or make any warranties as to the accuracy, timeliness, or suitability of the information in these podcasts. The Cardano Foundation has no responsibility or liability for: any claim attributable to errors, omissions, or other inaccuracies in these podcasts; or for any decision, other acts or omissions made in connection with your use of the information presented in these podcasts. Unless stated otherwise, reference to any specific product or entity does not constitute an endorsement or recommendation by the Cardano Foundation or by the participants in these podcasts.

This page may contain links to third-party sites that are not under the control of the Cardano Foundation. Access to such information does not imply association with, endorsement of, approval of, or recommendation by the Cardano Foundation of the site or its operators, and the Cardano Foundation is not responsible for the products, services, or other content hosted therein.

Transcript

SPEAKER_00 0:00

Welcome to Let's Talk Cardano, presented by the Cardano Foundation. In each episode, we delve deep into the transformative world of blockchain technology. Join us as we explore how blockchain is reshaping industries from finance to supply chain and talk with some of the key pioneers at the forefront of this revolution. All right here on Let's Talk Cardano.

SPEAKER_02 0:26

In this episode of Let's Talk Cardano, we speak with Douglas Heinzman about how digital trust infrastructure can support secure transactions and connect blockchain, identity, and AI systems at scale.

SPEAKER_01 0:38

Good morning, Douglas. Good morning. How are you? Thank you. Amazing. How are you?

SPEAKER_03 0:43

I'm great. Happy to be at the uh Cardano Summit here in Berlin.

SPEAKER_01 0:47

Yeah, it's the first day, our second session today. I'm excited to have you here.

SPEAKER_03 0:53

I'm excited to be here.

SPEAKER_01 0:54

Today we're gonna see you once more on stage. But first, you're here in my little studio private office. Um I'd love to talk about today about research, what you've been leading since a while. And specifically, also if you uh want to talk a little bit about what you will present later with the digital trust infrastructure, but let's see where we're going. Okay. Who are you? Who am I? Who is Douglas Heisman Heisman who I look in the mirror every morning and ask myself that question.

SPEAKER_03 1:31

I've I've been in the technology industry for a long time. I spent most of my career at IBM and most of that as a headquarter executive. So I was responsible for technical strategy for IBM Software Group worldwide. And I've been a management consultant and a VP of strategy and a COO. And I'm currently two things. I'm the what's called the chief catalyst at the Blockchain Research Institute BRI, which is a global think tank that helps uh governments and companies chart their Web3 and AI journeys. And I'm also the CEO of a company called Sincuro, which is a cognitive artificial intelligence company specializing in process automation. So that's kind of what I do. And I'm I'm from Toronto, Canada.

SPEAKER_01 2:16

Thank you for that intro. Directly what triggered me was the title Chief Catalyst. I think it's an exciting title. Yeah. And it sounds really curious. What are you the catalyzer of?

SPEAKER_03 2:29

Yeah, well, our view is that there is this inherent potential to create value. And that potential is uh doesn't achieve its possible outcome unless it's catalyzed, right? That that there's a lot of technologies and a lot of concepts and a lot of regulatory constraints. And people have to kind of navigate through that and figure out how to bring all these pieces together so they can chart a strategy forward. And the catalyst is someone that can help them understand how all these pieces fit together and help them form that strategic framework. Um, frameworks are important, right? Because frameworks give us a common language and a common set of ideas so that we can have intelligent conversations, so we can make good quality collective decisions. And a catalyst is something that helps bring all those pieces together so that business leaders and policymakers can have that intellectual framework within which they can make high-quality decisions.

SPEAKER_01 3:37

You're building bridges.

SPEAKER_03 3:39

Building bridges, yeah. Which is kind of coincidentally very much the topic of the research that we've been working on and that we're presenting here at the summit.

SPEAKER_01 3:50

Yeah, brilliant. Totally not intended, but so the title of your session is Building Bridges?

SPEAKER_03 3:57

Well, no, the title is introducing DTI, Digital Trust Infrastructure. And bridges are a form of infrastructure, right? Eras and societies and economies are built on top of infrastructure. Infrastructure is when a group of technologies reach a level of maturity and then they coalesce into some sort of whole that has, that is typically regulated, it has accepted standards, and its complexity is abstracted. And so that we can then create value on top of it, right? Think about road systems, think about rail systems and ports and electrical grids and telecommunication grids. With electricity, we just assume that there's this outlet in the wall that we can plug something into, and that's allowed us to innovate all kinds of amazing things, lights and cameras and cell phones and toasters. That's a huge amount of innovation that becomes possible because the infrastructure to support that innovation has been standardized and abstracted and formalized so much that that potential gets unleashed. And so the question that we've been asking ourselves is what is the infrastructure that will underpin the next economy? In much the same way that roads and rail systems underpinned the industrial age, and that telecommunications and the internet underpinned the first part of the information age. As we move forward, what is the infrastructure that will underpin the next economy? And that's the question that the Cardano Foundation came to us and asked us. And that was that's the research that we've been working on for most of the last five months or so.

SPEAKER_01 5:40

Cool. It beautifully coincidentally ties very much into the previous conversation I had where it was about law and policymaking. And we discussed that policymakers did look into how a car works before they decided how fast you can go, what a traffic light should do, why there is a right before-left rule before they designed all those traffic signs. So it was always function first and regulation then.

SPEAKER_03 6:16

Yeah, this is a there's always been a bit of a tension between the concept of the need for regulation to protect personal privacy, to protect public safety. So we have rules that allow us to effectively live together in a society. And it's important to have those rules. And at the same time, rules, when they're overapplied, can inhibit innovation because innovation loves experimentation. It likes, you know, it likes landscapes that aren't overly restricted because we're creating new value into new spaces that the regulations that were in place that govern the previous generations, the previous generations of technology, the previous social norms, the previous any number of things may in fact restrict the ability to do innovation. And so there's always been this conflict. And the trick is to figure out how to make sure that the regulation doesn't inhibit the innovation during that the rapid innovation phase. And as the technologies are maturing and coalescing, that the innovation, that the regulation then comes in to normalize and standardize them so that they can be adopted by a lot of people. So at some point, you need to come in and say there's a group of regulations that's going to allow mass adoption of the infrastructure. And so that timing is very important. And the other thing that's especially true of digital trust infrastructure, which is the argument that we're making that will be underpinning the next economy, is that you also want to embed into the fabric of this new infrastructure some of those compliance issues so that regulators, so that auditors, so that the people that are enforcing the rules that allow us to work productively together can easily do that without overly burdening the people that use the infrastructure. So, because this is an information infrastructure and a transactional infrastructure, we have the ability to embed a lot of that stuff into the foundation of the infrastructure, once again, to liberate the people that use the infrastructure. So they're not overly burdened. It will be much easier to deal with licensing and getting approvals and dealing with regulatory oversight and compliance issues because those elements are embedded into the fabric of the infrastructure itself.

SPEAKER_01 8:57

Is it defined time already for that type of standardization?

SPEAKER_03 9:01

That's a great question. Timing is everything in technology. And we have these magical moments in the evolution of technology for the last, you know, 10,000 years for that matter, where the discrete pieces of technology emerge and evolve on their own, and then they coalesce into something that we define as infrastructure. And then suddenly there is this step function where suddenly there's a massive amount of value creation because people have access to a new infrastructure, right? This happened once again with road systems, with rail systems, with electricity, with telecommunications, with the internet. And we think it's going to happen again. So the question of is this the right timing? In some ways it kind of has to be, because we so the so there's there's there's really two different parts of the answer to the question. One is, are the technologies, have they reached the level of maturity that says that they're ready to be coalesced and to be defined as a framework, as infrastructure? And then there's the question of, is society both ready for it and does it need it? And one of the really interesting things on that second part that a lot of people haven't really necessarily thought through is that economic growth has always been based on three different factors: ready and access to plentiful and inexpensive labor, access to plentiful and inexpensive capital, and factors of productivity. And because of the demographic profiles in most industrialized countries, access to inexpensive young creative labor is really problematic because our demographic uh pyramids are inverted. The capital that has largely subsidized economic growth for the last 30 years has largely been the retirement funds of the baby boomers, which is now being withdrawn and is going into travel and uh retirement and those sorts of things. And so that means that economic growth is really dependent on factors of productivity. We need to become radically more productive. We need to automate a lot more stuff, or else we're going to have a lot of trouble dealing with all the other stresses in the world. And so we need a new kind of infrastructure to support a new kind of economy that has, frankly, pretty radically new kinds of levels of productivity. So I hope desperately that that this convergence of this very fundamental societal need paired with the maturity of these underlying component technologies is in fact the dynamic that causes one of those moments that allows that stepwise function. It says that the world after the emergence of this infrastructure is quite different than the world that was before it.

SPEAKER_01 11:48

Now I understand why you said it has to be. Because of quite deep societal factor. They want to even call it issue of that inverted pyramid. Really, really good. Really, really good explanation of that. What about the maturity of the technology? Didn't we leave the playground yet? Or is there more to more sandcastles to build?

SPEAKER_03 12:15

Well, well, there's always more sandcastles to build. And this is especially true of the AI component, because in the model that we've built, the framework, there's a bunch of different levels, right? So maybe I should just quickly explain what those levels are and how they all and we can talk about the maturity of each of those levels. So the foundation of a digital trust infrastructure is a trust anchor, right? And it's the reality that that data and information are the basis of all transactions. But we need to have trust in that information, right? We have to have trust in the information that is the foundation of any transaction, then we have to have trust in the record of that transaction. And the reality is that those, the, the data elements will be stored in all kinds of different data repositories all over the place. But that we now have, because of blockchain, we now have the ability to create truth anchors. So it says that no matter where the data technically resides, and certainly some of it will reside on chain, but much of it, in fact, most of it won't be on chain. But we can create these trust anchors, these records of the timestamp that this piece of information at this time, on this, at this, in this place, in the context of this transaction, this is what the truth was. And that we can all agree on that. And the fact that we all have that common understanding of what a truth is gives us a lot of flexibility to do transactions without worrying about the trustworthiness of a counterparty, because it's it's kind of defined and and it's embedded in the trust, it's embedded in the actual network itself, in the foundation of the infrastructure.

SPEAKER_01 13:51

Before we go into layer two, it is, I mean, it's a framework and it and it's abstract non-purpose, but can you de-abstractize that that what you said about who are we creating the trust for? What specific application are you referring to when you consider those? Um Yeah.

SPEAKER_03 14:09

So the the the most obvious, the very first use case that everyone's familiar with, right, is is cryptocurrencies. And this is one of the most interesting things because a lot of what we nearly need to do is differentiate the what cryptocurrencies from blockchain, because they're not synonymous. Blockchain's an enabling technology that allows cryptocurrencies to happen. And the reason it does so is that cryptocurrencies have always suffered from the double spend problem, right? I can I can't just give you a copy of 10 euro because it means that the original 10 euro still exists and it devalues the nature of that medium of exchange. So you need to, you need to enforce digital scarcity. And the blockchain allows that transaction to happen so that when it comes out of my wallet and goes into your wallet, that happens at exactly the same time. And the transaction is ledgered so that everyone else in the world knows that you now have the 10 euro and I don't have it anymore. And that really is a fundamental truth anchor right there. That's what that fundamentally is. Now, if we apply that technology to lots of other kinds of things, right? So think of supply chains or cold chains, removing pharmaceuticals or vaccines around the world. And who has the custodianship of a particular batch of vaccines when there's a temperature excursion and the vaccines get spoiled? And who bears the responsibility of that? Um, and if we have advanced knowledge of that, how do we then insert new product into the supply chain to replace it? If we're worried about ethical sourcing, whether it's it's materials coming from areas that are being deforested or that involve child labor or conflict, that that we have confidence that those goods are coming from places. We understand the provenance of those goods and services. These are all fundamental use cases that are then incorporated into much more sophisticated composite use cases.

SPEAKER_01 16:03

Yeah. Okay.

SPEAKER_03 16:05

And if we know the truth about those things, it just removes so much friction. And that kind of takes us to the second level, right? So once you get beyond the trusted data fabric piece, you get into the area of identification. Identification means a lot of things, right? So there's core identity, which is who are you, right? What is your essence? And when you when you define yourself as someone else, how do they have confidence that you are who you say you are? Then there's kind of another shell that goes beyond that, that is really about what are your attributes, right? What is your what is your education? What is your licensing? What is your credit score? And what are the transactions that make up that credit score? So there's that, these are attributes that are are structural and kind of fundamental to defining your extended personality. And then you've got a social reputation, right, which has to deal with, you know, how trustworthy you have been in the past, your social media expression and interactions that define these kind of softer elements of your identification. And then there's this operational space. And that the operational space is something we really call the identic layer. And that is because we're at the advent of a new era where we will have intelligent agents that operate on their own on behalf of a company or even on behalf of an individual and transact on our behalf and advocate on our behalf and make investment decisions on our behalf. And we need to be able to express the ownership of those agents. We need to be able to express what technology is backing those agents, what they've been trained on, right? When they were last certified, who oversaw them, those sorts of things so that they can navigate the network, interact with DAOs and banks and all these different entities and transact with high degrees of confidence. And so that's the outer shell. Beyond those identity elements, we have verifiable credentials, which says that I go for a job and I need to explain why I'm qualified to do the job. But this is also true of an IoT device, right? That I am an IoT device and I'm representing that I measure these kinds of things and that I am in this place. And so that the information that comes out of me has a contact. And as well as agents, right? Agents will have different kinds of credentials associated with them that give people confidence that they can interact with them. And then beyond that, we'll have what are called presentations. And presentations are really powerful construct because they allow me to express a capability or a state without having to go into a lot of detail. Right. I mean, the thing that we can all relate to is we go into a bar or or we we have some sort of interaction that there is a minimum age. All we have to do is say, I qualify. I'm over that age. You don't need to know when I was born or where I was born or who my parents were, all that kind of information. It's just a state, right? I'm allowed to vote. Digital passport for a product, right? Coming across a border. This product complies to these regulations. It hasn't been sourced unethically, right? It's got this carbon content, which is below this level, green. Go ahead. I don't need all the details because the details are encapsulated inside that presentation. And all of the certifications and all the proofs that demonstrate that I have this capability, they're all just that's all I need to do is interact with it. Once again, reducing huge amounts of friction out the back end. So identity is the next layer. And then on top of that, you've got you've got certified data exchanges and authoritative registries, and that really has to do with interoperability issues. And then above that, you've actually got something called basic universal. Oh, sorry, above that you've got uh payment rails, right? Which we've kind of touched a little bit with cryptocurrencies, but the advent of centralized digital currencies, central bank digital currencies, as well as uh stable coins that are pegged to fiat currencies, combined with various different um uh crypt uh payment mechanisms and all the various different financial tools that have been evolving. You ask the question about maturity. We've had a good 15 years of maturity of a lot of these tools and a lot of experimentation. Some things worked and some things didn't work, but it's remarkable how much advanced technology has emerged in the last 15 years around this. So digital payment rails, the next layer, and then above that is the universal basic intelligence. And AI plays a twofold role in the in the infrastructure. One is that it helps support the the veracity and the integrity of the infrastructure itself. It's monitoring for cybersecurity attacks, and it is doing kind of prediction about the the structure of transactions to figure out if things are, you know, should be flagged as worrisome, that kind of thing. But more importantly, it becomes a services layer where we strongly suspect, and we've already seen evidence that there's companies doing this, but that that actual jurisdictions will want to have basic intelligence that is trained on the jurisdictional and legal elements of a particular country or region, the the business norms, the cultural norms, the way of communication, the native language, that these are distinctive qualities of participating in commerce and societal exchanges within a jurisdiction, and that everyone is going to want to have access. To those things. A small business will want to understand what the local regulations are and how to get the right licensing and how to find the right partners and the right customers and how to optimize sales and marketing for that particular audience. That intelligence, the ability to just query a system and find the answers that are specifically relevant to them in their situation is incredibly powerful. And so that's the top layer of this infrastructure. And then there's lots of pieces that allow them to all work together and tie together. Yeah.

SPEAKER_01 22:34

Sounds like this massive governmental data lake. In data science, you often talk about data lakes, which just dump all the data that you want to somehow have around into there.

SPEAKER_03 22:46

Yeah, I think the in the big data era, right, during the mid-2010s, there was very much this idea that said, we got lots and lots of data, let's throw into a data lake. And of course, then as AI emerged and said, let's train AI on those big data lakes. As amazing as some of the big frontier models are, and training on the corpus of what's out there in the internet and Wikipedia and all these sorts of things, or X posts. Vastly more information that we have not come anywhere close to tapping. And for good reason, right? These companies and hospitals and governments, they don't want that data to be leaking all over the place, either for privacy reasons or for company secret and proprietary information reasons. So I think the reality is that the training sets for this AI and the training of the AI itself will become increasingly decentralized. It'll become much more federated, where whether it's the training of the models themselves or it is the real-time augmentation of those models, the contextualization of the queries going against those models, will be derived from various different smaller distributed centers, right? That they'll be within the context of a company, the context of the country. And so that'll be a lot of the challenge going forward is figuring out how to effectively federate that information while protecting the privacy and the competitive information.

SPEAKER_01 24:35

You know, I'd argue that it wouldn't even be necessary to federate the training of that data. Because there will always be and there should always be privacy or or the knowledge of the data is valuable within the company. And if there is value to somebody else, then there is worth economic worth for that other person or for that other agent to be part of that. If we're now considering an agent that knows the inside data and another agent that knows that that agent knows, but not the data itself, there could be just a transaction, and we have agents that are representing the data behind it.

SPEAKER_03 25:16

Yeah, that's that's an interesting view. And I think you're right to a very large extent that we will see these swarms of agents that some agents will present themselves as tools, right, that other agents can query. And the agent then goes into the back and does some work. The agents themselves, at some point, need to, well, they'll probably be composed of a bunch of different underlying technologies. Some of them may use recursive techniques and reinforcement learning techniques, but there will be, there will be an awful lot of them that are based on models. And whether they're frontier models that are, you know, these big uh Grok and Anthropic and OpenAI and these sort of big frontier models, or whether they're based on small language models or domain-specific models that are deeply trained on a particular company's data sets and incorporate the intelligence of those data sets. I think that at some point those agents themselves will, first of all, have to be informed and in many cases trained by those models. And they'll also make use of those models, right? They will invoke those models as tools. So how the structure evolves, whether the interphase to all that information is moderated by agents, is I think a level of detail that will be optimized in different situations in different kinds of ways and will evolve. But for most people, that will be abstracted very low in the way down in the stack, right? We'll be interacting with the systems and asking questions, interacting and transacting, and all that complexity will be at a much lower level.

SPEAKER_01 27:05

I 100% agree. And it is quite a detailed, detailed technical solution question. But to to let you know my thoughts on why I think that way, you said that it will be necessary to decentralize training of data.

SPEAKER_04 27:20

Yeah.

SPEAKER_01 27:21

And I have I strongly believe that it's fundamentally not possible because of how training works today. Training a new model on data needs to happen in a as close as possible together because of something interconnect. That's something called interconnect, which is the thing that the matrix multiple multiplications happen between the chips to create those nodes and to create all those values in the model. If that were to be physically decentralized, the lag or the delay between the compute centers where that can actually happen would be way too large to make the already extensive energy use and the already extensive time that it takes to create an even better frontier model, because that conversation is really only about AGI type pushing models. If we're talking about specific ones, then it will not be necessary. But the decentralizing the training part for that is, I think, not solvable.

SPEAKER_03 28:34

Well, I think that's a very conventional view. And I think I myself would have argued very strongly for that point of view not too long ago. And every time I think that I've figured it all out and that, you know, there are just some rules of physics, right? That interconnects energy consumption and heat generation means that it has to be done this way. And the economics mean that it has to be done this way. Something new happens, right? We have that deep seek moment that says that, you know, radical distillation changes the economics completely. Oh, wow, right. And just recently, the the unveiling of the Samsung 7M model, which is a tiny recursive model, which has achieved the Arc AGI benchmark that exceeds the Gemini 2.5 considerably. And it's a tiny model, right? It's 7M. It's but it but it uses massive recursion.

SPEAKER_01 29:30

It's kind of the difference between a I thought I thought 7M is just a name. No. 7M. 7M. What?

SPEAKER_03 29:36

Yeah, exactly. But it's a recursive model. So it does something very simple. It kind of guesses, makes the best estimate what an answer is, and then it rethinks the problem over and over and over again. Think of it like the difference between a CISC architecture, a complex instruction set architecture, which has a lot of complexity in it, and a RISC architecture that just has very simple instructions but does them over and over and over and over again, really, really fast. Well, of those two models, the RISC architecture is the one that succeeded, right? It's the one that's in our phones, right? And it's it's in our new laptops because it's a lot more energy efficient than the Cisco architecture. It may be now that the route to AGI doesn't actually run through LLMs, because after all, at the end of the day, language is a relatively inefficient way to describe symbolic logic, right? And it has some serious limitations and it's it's a lot of has a lot of overhead associated with it. It may very well be that another model, like massive recursion, is in fact a much more efficient route towards AGI. So I think we need to kind of suspend our beliefs in the the hard paradigms that we've we've been living with. Now, going back to your your federation, I think it's quite true, at least for the foreseeable future, that we will need massive data centers with very fast interconnects and big GPUs and big racks that are just churning through and generating these multiple hundred of billion parameter models. And there are going to be a few of those foundation models. Then we're going to have a bunch of models that are going to be using radical distillation, that are going to be condensing those things and using communities of experts to do really interesting logic and decision making that use many fewer resources. And then we're going to have these tiny recursive models that are going to be doing a lot of edge computing. I think we're doing a lot of AI, especially in the inference part of the equation, on the edge, right? In in our cars these days, when we're using automated driving, right, or autonomous driving, we're not connected to a network into some big data center that's running a whole bunch of GPUs with fast interconnects. We're processing that decision making on the edge.

SPEAKER_04 32:02

Yeah.

SPEAKER_03 32:03

And that's getting better and better and better all the time. So when I think of federation, I think that there will be those big foundation models. Absolutely. But increasingly, there's going to be market plates of data where companies are going to put their data sets up. And you'll be able to come in and train on various different data sets from a data marketplace. And you're going to have various different distributed models that are trained in this federated model because the trade-off of having big multiple hundred parameter models or 100 billion parameter models can be offset by the advantage of having access to very domain specific pieces of information that are informing a much smaller model. And we'll have lots and lots of those things all over the place. And an inference time, to the point you were making, we'll likely have, and perhaps it is an agent interface, we'll have agents that basically call tools or other agents that will inquire to various different distributed smaller models and aggregate and do best, best answer analysis from all those different things and pool them all together to provide you with the answer. So it's always a trade-off. The idea that we take the entire world's data and put it in one place and just train a multi-hundred trillion parameter model on everything, that doesn't work either, right? We don't have enough energy on the planet. It would be and it would be incredibly expensive and have vast amounts of latency to do the vast number of regular tasks that we would want such a model to do. We're going to have all kinds of different models, some of them tiny, some of them larger, some of them in between. They're going to be all over the place. And we're going to need to coordinate them.

SPEAKER_01 33:55

It sounds a bit as if you're saying AGI is a myth.

SPEAKER_03 33:58

I don't think I said that. I think it's probably more true that the definition keeps on changing, right? And that no one can agree on what it means, right? It's one of those things that perhaps you'll know it when you see it. I suspect that the average person from 15 years ago, if you had sat them down in front of ChatGPT, would have said that's AGI.

SPEAKER_01 34:21

That's AGI, yeah.

SPEAKER_03 34:22

Right. But as soon as we have ChatGPT, we're saying, well, it's not really AGI, right? And so we'll keep on moving the bar because we don't really know what it is. No one really agrees what it is.

SPEAKER_01 34:33

We consider this personified thing, whether it's a chat box or whether it's uh some server. Doesn't really matter, but we think that is AGI. What you were painting to me sounded as if that could be AGI. This whole network of agents and models that work together in order to like solve or answer any question there is.

SPEAKER_03 35:04

I think that technically speaking, it would be probably more accurate to describe that as a genetic workflow, right? The idea that you've got a collection of agents that work with each other and that call specific tooling to do specific things, and that there's some agents that have managerial and strategic oversight and have goal definition as part of their agency, and that call other agents that may have less amount of agencies. They may be very specific things that just do one thing all the time, really. You know, they they ask what the weather is, they ask what a postal code is, you know, they do really simple sort of things, data lookup sort of stuff, that collectively they may represent something that we could give an AGI label to, because they do so many discrete different pieces of the puzzle that if they're coordinated, they act in that way. In much the same way that a company has some marketing experts and some sales experts and some research experts, and that collectively they form a company intelligence that interacts with the outside world with a collective competence that no one individual has. And I don't necessarily think we should get hung up on those labels. You know, yes, I think they're the risk of us moving towards a super intelligence, an intelligence that can design an intelligence smarter than it is, and so on and so on. It's science, it has been science fiction. We're starting to realize that we may actually have the science to go there, and that that's something that we should worry about. Absolutely. I am by nature a bit of a techno optimist. I think it is true that we should have some real discussions about where all this leads and try to put in place some of the guardrails that are reasonable. And this kind of goes full circle to where we started. That as with a lot of these technologies, we have this tension between regulations that safeguard the welfare of society and of individuals, and the resistance of limiting innovation, because innovation can create value and make all of our lives better and increase productivity. And where do we find the right balance between those two desires? And when does, you know, how do we let this go, the the innovation side go really, really fast? How do we know what the right timing is to come in with the just a second before we take this next step? We need to put these rules around it. And that's gonna be a lot of the art.

SPEAKER_01 38:01

Yeah. If we had two, three hours, I'd I'd have like tons of more questions about language abstraction of information. That's one of my recent pet peefs and AGI and so on. But I do, as you also just probably hinted on, and I want to bring it back to blockchain and your research, but it was a great excursion. If you consider, if if you think back of when Cardano Foundation reached out to to contract you with said research. Commission, yeah. Commission commission it. What has been your favorite aha moment?

SPEAKER_03 38:41

Wow. That's a great question. I was quite lucky to work with a number of people from the Cardano Foundation that that helped me really work through a lot of the technical aspects and some of the art of the possible questions. But I also was able to interview industry experts, policymakers, leaders, academics from all over the world, from, I forget, a large number of countries. And those discussions were really fascinating. There were many aha moments that when I was I was talking to the CTO of MD Anderson and he was talking about the fact that we need to move beyond the the privacy regulations. In the US, it's called HIPAA, but there's similar regulations in other geographies that in this new era we need to move towards a consent model, right? And that those are two very different kinds of worlds. And the transition from one world to the other world is has all kinds of challenges associated with it. I was talking to the CEO of Autonomous in Saudi Arabia, and uh he he really kept on hammering me over the head about the importance of public-private partnerships, right? That you can build all the best technology in the world, and it can be completely inevitable. But unless you really work really, really hard on the public-private infrastructure, the public-private partnership, and figure out who is going to do what and where the funding is going to come from and where the value is going to flow and how the regulations are gonna work, that that was a really important, that that was an essential component that as a fundamentally as a technologist, that I hadn't really appreciated until I heard his passion about this, which was just complete pragmatism. The technology is great, but if you don't figure this out, you're in trouble. And then I suppose the other thing is that governance is the flywheel. That, you know, we are in some ways, the technology side of the equation is kind of the easy part, right? It's going to happen because there's just a lot of smart people and they talk to each other and they come up with smart ideas. The governance part of the equation about how we're going to plug all this stuff together and make it work and make it predictable and make it reliable and make it safe, those are really challenging problems because, in many ways, they're human problems, right? The human problems are always much more difficult to deal with. And there's lots of examples out there of people tackling this problem. And so I've a lot of optimism that we'll figure it out. You know, there were so many, I think if I had to really sum it up, the big light bulb was the transition about the way we need to think about things. Whereas the world we've lived in is really about information generation, information storage, and information transmission. Right? That's kind of been the fundamental foundation of the first part of the information age. We're transitioning to a model that is much more a request and verify, right? That I will, instead of you sending me messages through a messaging interface between your data set and my data set as a means of transacting, I'm going to run around requesting access to information. And in response to that request, various different credentials or presentations of capabilities will be immediately available to me. And I'll be able to have an enormous amount of trust that that presentation, that those credentials are in fact accurate. And that really changes how much friction there is. Because in the capture, store, and transmission kind of world, I still run into this credibility problem, right? In Europe, we have this enormous problem where people are fraudulently claiming signatory rights for a company and transacting with other companies and only to find out that it's it's a fraud, right? Either it's a malicious fraud or it's uh an accidental fraud, because in fact they don't actually have those signatory rights. But they're they're representing that they do. And because this is happening, every time you interact with someone, I have a trust deficit. And I have to look into it. I've got to do research, I've got to get a third party to validate them or to you know claim on their behalf and represent that, yes, in fact, these guys are good guys and you can you can transact with them. And and the consequences of those frauds are terrible. Trying to unwind it and just deal with the legalities and of course the lost the lost money and the lost time is is really debilitating. That that that simple transition of the way we think about the world to one which is a a a request, a query and verify sort of world is I think much more profound than people realize. And it's going to allow us to design completely new kinds of systems that that have much, much less friction in them because trust. Is now just an inherent part of an enabling infrastructure. I think that's the big idea.

SPEAKER_01 44:07

You you keep saying we and us and we we gotta figure it out. It's a human problem. And 100% agree and yes, but each individual within the we has their own incentives. And even we don't even have to go as far as a digital trust infrastructure future that is standardized, we just have to look at proof of concepts today. If you go to the CoinMarketCap, you see 200 different layer one blockchains that essentially do the same thing. And depending on where the incentive of said person that is driving a proof of concept, where the incentives are, they may choose one technology over another over completely arbitrary things that have nothing to do with a collective we or a standardization or a collective better future if we were to agree on one chain. The question I want to ask is like partially, do you agree? And and secondly, is how would human incentives be solvable in that in that statement that you're raising, that we are going to a better place, that we are leveling up our infrastructure?

SPEAKER_03 45:33

Yeah, there's actually a variation on that that uh is called the consortium problem. And we've seen that problem before, right? I was a huge believer that blockchain was going to transform supply chains because it just makes so much logical sense that if you've got a common ledger across all the supply chain, that everyone in the supply chain has visibility to where stuff is and what its status is, that the entire supply chain can work a lot more efficiently. Okay. Makes total logical sense. And yet, if we're honest about it, it really hasn't changed the world in that way yet. And we saw examples like TradeLens, right? The Mersk IBM consortium of let's get everyone together and and you know, build everyone on the same blockchain and we're going to change shipping and logistics and supply chain. And it really didn't work because it was to the problem you were on. It's difficult to align the interests of a lot of different people that have different agendas. And some of them may just shrug their shoulders and say, I just I've always done it this way. I want to continue doing it this way. And this is just it's work, and I don't need the work, and I don't get enough benefit, enough value doesn't flow to me to be for it to be worth my time to bother with it. And that dynamic, that instinct has been consistent throughout most of human history because it's just kind of part of the human condition. I've always believed that that is true until we hit a value threshold. And the value threshold is when things coalesce enough, are standard enough, that have good enough governance around them, that there's regulatory frameworks that allow consequences of excursions to be controlled. That you hit these value propositions, sorry, these value um thresholds, that adoption has always been spotty. There's some enthusiasts here and there that do things in a certain way, but it doesn't quite coalesce until there's so much value in doing it in a new way that everyone suddenly does it in that way. Right. So, you know, we we've seen these cases, and the the time of adoption has, for all these new technologies, has been increasingly getting smaller and smaller and smaller because we find those special moments where suddenly there's a value inflection and suddenly everyone starts doing it in a new kind of way because it's so much, there's so much value in doing it that way. And I think there's a few things. So in the case of TradeLens, I always, my my analysis of why TradeLens failed was that they were trying to solve a Web3 problem with a Web 2 business model. Right? I think that was the fundamental problem. And that there are examples of other trade-enabling networks, blockchain-based networks, that are much better at figuring out how, in fact, to distribute the value and deal with that kind of transparency. I think that there are public digital infrastructure initiatives being driven by the United Nations and going into places like Estonia and Finland and India, and there's lots of geographies around the world that are adopting DPI, digital public infrastructure, which in some ways is a simpler version of DTI. Digital trust infrastructure is a much bigger framework than DTI. But DPI, digital public infrastructure, is about delivering governmental services to people much more efficiently. And they need a digital identity to do that, right? Which is kind of foundational to DPI. They need digital payment rails to do that, which is another piece of DPI. And there, governments just declare them, right? So in Switzerland, they just passed a new referendum that they're moving to a new DPI infrastructure with a new digital identity for the whole country. And lots and lots of countries are doing the same sort of thing. Poland's got some really interesting work going on. And so in those kinds of cases, it's effectively a mandate that says, listen, if you want to interact with governmental services, we're going to make it a lot less expensive and a lot easier for you to use this technology to do it. Sure, you can do it the old way, but it's more expensive and it takes a lot of time. So suddenly everyone does it this new way. Okay, well, that's a first step. You know, we see efforts in Europe with things like EPSI where we're trying to normalize credentials so that people that have certain certifications and licenses in one geography can move seamlessly from geography to geography with those credentials intact so that they're recognized. So we're seeing these initiatives. And once again, the value becomes so obvious so quickly that you start to see mass adoption. And the governments that are deploying these infrastructures are explicitly, you know, putting adoption campaigns in place to make the value, the rationale for moving to this new infrastructure very obvious to people, so that they say, well, why would I do it any other way? And I think we're going to see similar techniques and similar patterns emerge in the broader DTI ecosystem enablement space, where ecosystems will use some of those same lessons learned and say, you know, listen, if you want to interact with us, right, either we're big enough because we're Walmart and we're a keystone in our ecosystem to say, listen, everyone has to do it this way, or else you don't participate in our ecosystem. Or that the ecosystem says, listen, everyone can participate. But those of you who use this infrastructure have all of these advantages, right? Suddenly they're using all this AI magic stuff and they're transacting and moving money around super efficiently, and they have all these presentations so that they can move across borders, right? You get a border guard, you have two different, two different lane ways, right? One for people that are doing it the old way, and one where the entire load inside the truck has been pre-certified. Everything's got all of its ethical sourcing stuff and its carbon stuff and its import stuff, and everything's all been pre-checked, and it shows up and a transponder identifies it to the border crossing, and the border crossing guard screen suddenly turns green and go ahead. Wow. If you were a truck driver, why would you want to do it the old way?

SPEAKER_04 51:55

Right?

SPEAKER_03 51:55

It's just so ridiculously better to do it the new way. And I think we're going to see a lot of those value thresholds as the tech, as the infrastructure coalesces, that those value thresholds will be hit and the the advantages become so obvious that mass adoption um happens pretty quickly.

SPEAKER_01 52:17

Like tipping points towards um hyperbluctionization term.

SPEAKER_03 52:25

Well, the other thing is that I remember back in the mid-90s, there were a bunch of products. We used to have software in boxes back in the good old days. And and a lot of the boxes had these bright yellow starbursts on that said featuring TCP IP, right? The technology was a a feature of the product. But of course, as the as the technology evolved and emerged and became fundamental, we took it off because it's kind of boring. The reality is that you know we use TCP IP in everything, but we don't think about it. And I've always thought that technologies really succeed when we stop thinking about it. Right? Blockchain will have really arrived when we stop using the word blockchain, because it's, of course, it's just it's everywhere. It's a truth anchor. I mean, even in the research we've done, of course, we use the word blockchain as we're describing various different pieces of technology. But fundamentally, we really talk about it as a trusted data fabric, which is a more comprehensive term that has ledgers under the covers that are providing that truth anchoring, but it recognizes that data's in a lot of places and that there's going to be trusted registries and there's going to be authoritative registries. And that's just the reality. We're going to have a hybrid data space, but underpinning it is going to be this trust anchor layer. But collectively, from an architectural standpoint, we need to just think about it as a trusted data fabric. And then we can build lots and lots of stuff on it.

SPEAKER_01 54:02

Well, I think it does matter. Actual blockchain infrastructure beneath it?

SPEAKER_03 54:06

I think it does matter. And I think we will have maybe not your many thousands, but we'll have lots of them. And some of them will be optimized for different things. There's going to be some of them that are going to be optimized for ledgering sensor input from IoT devices. There's going to be some of them that are optimized for building reputational marketplaces for AI agents so that I can go to a marketplace of AI agents, a swarm, and pose questions or ask for work to be done. And they'll, amongst other things, use a reputational ability to for the various different modules to bid or to stake.

SPEAKER_04 54:45

Right.

SPEAKER_03 54:46

And so we'll have a whole token economics around how these marketplaces work. So I think we'll have lots of these. And some of them will, you know, just underpin. I mean, Bitcoin will probably continue just underpin a crypto bearer asset, right? Because it's a digital goal, but it's not a terribly efficient or effective blockchain to do a lot of stuff because it's incredibly expensive. You know, Cardano has a lot of attributes and a lot of uh of great ecosystem participants generating a lot of value. And it's doing tremendous work in areas like ESGs and carbon tracing and those sorts of things. And so they'll be used for those things. And so we'll have a bunch of those different ones and you know, Hyperledger will be used in, you know, pick something and in uh in healthcare for you know uh campuses of hospitals, those sorts of things. So there's gonna be lots of these things. The trick is going to be developing the interoperability, right? How do we get from one place to the other, right? I checked into my hotel room and all of my plugs have you know kind of slots and a and a and a grounding bar, right? And everything here are little round holes and they don't work together, but we have adapters. So I can put an adapter on mine and plug it into yours, and it works. And so we're going to have those kinds of adapters and interoperability standards.

SPEAKER_01 56:15

It's a mass analogy. Yeah, for multi-blockchain, for multi-blockchain world. Dr.

SPEAKER_03 56:22

And there, by the way, there will be normalization. It used to be that between France and Spain, the the rail gauges were different spacings, right? And some rail gauges can support different amounts of weight. And you'd get to the end of yours and you had to unload your train, walk it over and load it onto a new train and go off. And at some point, people said, Well, that's crazy. Let's just make one rail and put a standard across the entire system that says it should support this amount of weight, and then things move with less friction. So there will be those kinds of situations as well.

SPEAKER_01 56:54

But who or how is a decision made of whose earth is being torn up? Torn up and then new tracks are being built in because it it needs to be either France or Spain to back off their own system.

SPEAKER_03 57:09

Yeah, or a third one comes in, right?

SPEAKER_01 57:11

Or a third one comes in, which is the same as with blockchains. There's so much value in each of these ecosystems. Yeah, you may have now the larger cake and you may have the tipping point now, but we don't continue fighting until we are actually I think these things happen organically.

SPEAKER_03 57:28

There was a period back in the I don't know, 70s or 80s where the United Nations was trying to figure come up with a new language that we could all use to talk to each other with less friction. Yeah, I think it's called Esperanzo, right? So it was it was a derivative of Spanish with a bunch of extra stuff thrown in. And the idea was that everyone would learn this new language, and it would be the common thing that everyone used. Of course, it didn't work because no one spoke Esperanto, and there was no economy and no business that was built up around this newly invented language. And and English became that basic language that's pretty much everyone's second language. And that was largely because of just the economic force and the cultural force of the English-speaking nations at that point in historical development. I think that there's going to be some similar sort of dynamics where, in the case of France and Spain, there is more value on one side of the border than the other side of the border, and there would be more cost in retooling one side. And if the various different participants basically recognize the advantage of doing it in a common way, then the one that has the more powerful position and has more to lose because it's generating more value by changing, we'll talk to the other one and say, Well, you've got to change. And let's say, well, wait a second, it's cost us. Okay, I'll tell you what, we'll share some of our value to incentivize you to do it this way, because at the end of the day, we're both going to benefit from this. So we're going to see some of that same sort of thing happen. The better that in the case of the railways, you're right. There's very real cost. We have to dig up stuff and bring machinery in a lot of labor, and it's very expensive. In the case of some of these digital entities, it isn't as much cost, right? These things which we're not talking about lots of laborers and lots of backhoes and stuff.

SPEAKER_01 59:24

Actual cost, no, but incentives.

SPEAKER_03 59:26

Yeah, no, I think that's exactly. I think that's right. But I think that's also where some of the token economics comes in, right? That we can use the efficiency of value distribution to incentivize various different people so that they behave in pursuant of the collective good. And I think that's one of the great advantages of this new generation of infrastructure. Like all those other infrastructures we talked about before: roads, rails, telecommunication grids, electrical grids, they none of them had the advantage that inherent in the infrastructure was this ability to very efficiently move value around. Value was always a side channel to all those other infrastructures. This is a new kind of infrastructure that has value exchange as part of its inherent nature. And that gives it huge advantage because you're absolutely right. This all has to do with incentive alignment and getting people to kind of think the same way. And there's no more efficient way of doing that than saying, listen, world one, this is how much you have at the end of the day, world two, this is how much you have. Let's do this one. And we can figure out how to move all the value around really quickly and really efficiently. So if that logic holds, the time that it took to adopt into mass adoption, all those other kinds of infrastructures throughout history, I hope that it'll be a lot faster because, first of all, it's digital technology. But second of all, to your point, the ability to align people's interests in the pursuit of collective good will be dramatically faster and easier than it has been in previous infrastructure incarnations. I hope so. Well, see, it's a great paper. I hope uh uh I hope that it's valuable and I hope that that well, I hope everyone reads it and and that it's becomes a useful framework on how to think about this stuff and how to engage in productive discussions, both with business people as well as with policymakers. Yeah. Uh and and I I'm very grateful for to the the Cardano Foundation for supporting the research. And and and once again, we hope it'll just be very impactful.

SPEAKER_04 1:01:48

Yeah.

SPEAKER_01 1:01:49

I'm sure those that listen today to your talk will be grateful too. I'm grateful for having you here. Pleasure. It was a wonderful discussion. I really wish we had like another three and a half hours. Well, we'll come back.

SPEAKER_03 1:02:04

But I'll have you on my podcast. We could, yeah, we could really continue it there.

SPEAKER_01 1:02:09

Perfect. Pleasure. Thank you, Douglas. Take care.

SPEAKER_00 1:02:13

Thank you for joining us on Let's Talk Cardano. For more insights and deep dives into the world of blockchain. Don't forget to subscribe and reach us at CardanoFoundation.org, where you'll find extra resources and content on all things blockchain. Leave us a review wherever you enjoy podcasts. Follow us on social media, and stay tuned for more coming soon.