Podcast Transcript

Smart Tokens vs REST APIs: TokenScript Explained

Mathew Sweezey Former co-founder, Salesforce Web3 Studio; Smart Token Labs advisor (at recording)

In this interview, published in August 2023, Mathew Sweezey argues that tokens can deliver portable services, not just represent assets. His car-and-insurance example leads to a practical question from Kevin Riedl: why use smart tokens rather than agree on a REST API? They discuss TokenScript, off-chain data, attestations and metadata profiling. The transcript preserves the argument and its challenges; the sourced answers distinguish a token interface from automatic interoperability or privacy guarantees.

  • Published 02 Aug 2023
  • Runtime 24:10
  • Operator
  • Recorded in English
Smart Tokens vs REST APIs: TokenScript Explained: podcast episode with Mathew Sweezey, Former co-founder, Salesforce Web3 Studio; Smart Token Labs advisor (at recording)
// The short version
  • At 02:28, Sweezey reframes tokenization as tokens acting as applications. At 03:58, he illustrates a car token that exposes vehicle and insurance functions inside another service. The example explains portability and composition, not a confirmed Uber or Apple Wallet integration.
  • At 06:48, he argues that bilateral API integrations become difficult to maintain at scale. At 12:12, Riedl asks why firms cannot standardize a REST API instead. The interview proposes a token-centered interface; it does not demonstrate that APIs or common business semantics disappear.
  • At 13:00, Sweezey introduces Smart Layer and TokenScript. The TokenScript framework documentation describes a runtime and signed files that expose token functions, including off-chain operations. A receiving application still needs compatible support.
  • At 13:24, he describes a contract referring to an external script. ERC-5169 defines scriptURI for that purpose and discusses script integrity and server-API trust. Sweezey names the proposal as EIP-6159, hedging with “I believe”; the described mechanism matches ERC-5169. A script reference does not mean arbitrary code executes on-chain.
  • At 17:53, Riedl challenges whether public metadata can support unwanted profiling. Sweezey proposes off-chain data and limited disclosure. Neither makes a system automatically compliant. CNIL’s blockchain analysis calls for a specific assessment of personal-data processing; the interview’s blanket GDPR claims remain the speaker’s claims.

Introduction and Smart Token Labs

Watch this part · 0:00

Jingle 0:02 Wavect, the Web3 software company that understands what you want.

Kevin 0:09 Hi everyone, welcome at Wavect. Today with Mathew Sweezey, who is the former co-founder of the Salesforce Web3 Studio. He’s also currently a Web3 investor and advisor for a variety of companies. And today we will talk about why tokenization is actually a game changer. So thank you, Mathew, for being here.

Kevin 0:32 It’s really a pleasure talking with you. And yeah, starting out, I want to give you the opportunity to let us know what you’re currently working on, maybe how your day looks like and yeah, whatever you want to tell us about yourself.

Mathew 0:46 Oh, cool. Thanks, Kevin. It’s great to be here. Yeah, so I spend most of my day working on what I believe is the future of the internet, which is what I believe is called tokenization. I have a very different idea of tokenization. I’m currently advising a company called Smart Token Labs, who is working on a new protocol, helping kind of take that to market, called Smart Layer, which is really where we’re looking at the missing link between Web2 and Web3, being able to solve existing Web2 problems, such as integration and data sharing, with Web3 utilities.

What are tokens as applications? The car-and-insurance example

Watch this part · 1:15

Kevin 1:15 Okay. Before we get into that, there might be some viewers that have possibly never heard about tokenization. Could you maybe briefly explain what it is in as simple as possible?

Mathew 1:30 Sure. As simple as possible, tokenization is really just using a token as a new primitive of the internet. And tokens are not a new thing. They’ve been around a long time. So the majority of us know tokens through the term of essentially a currency like a Bitcoin or Ethereum. And that is a token that is a fungible token, meaning that it’s essentially a currency.

Mathew 1:50 It’s a digital currency backed up by the blockchain. You then have another type of token, which is another type of asset or ownership token, which is a non-fungible token, which just means each of those tokens is unique. So it’s not like a currency, it’s more like a trading card. And so those are the two types of core tokens that we know about.

Mathew 2:09 Now both of those are classified as asset classes, right? They’re usually to define ownership of something. They give you some basic rights and you can take them places and say, hey, listen, I have this, or you, if it’s a currency, you can spend it. Those are too simplistic and what most people know of and think of as tokenization.

Mathew 2:27 When I say tokenization, I mean something very different. I mean literally a new way that we connect the internet, and it’s not just currency and it’s not just proving ownership that you have something. It’s really tokens as applications. Now follow this logic and you can understand. If I want you to think about a business service or a function, right, how does one party deliver something to another party?

Mathew 2:52 Right, so let’s go back to when we first got the internet. We get the internet and we deliver business services first off through websites. The first thing you could do for a website was get information. Then we started to deliver service, meaning you would go to a company’s website to interact with the company directly, whether that was buy a product, configure a product, get a service directly, right?

Mathew 3:12 The next wave became applications. We then moved those services to applications because it was an easier way to deliver those services to the end consumer. Most all businesses have a website, most all businesses have an app because those are the easiest ways to deliver those services. Now if we think of the next wave, the next wave is tokenization of service.

Mathew 3:33 Now this is where it gets really tricky, so let me kind of pull this apart really quickly. In the first two examples, the website and the application, those are a central repository where the individual goes to deliver the service, centralized. If we’re to decentralize that, essentially make those services composable and portable, that’s called tokenization.

Mathew 3:52 Now, that’s a crazy concept. Let me give you an example of what that would even look like and be. Let’s say that you are a car manufacturer, and every time you create a car, you also create a car ownership token. This is not just to prove that you own the car. This is literally the functionality of the car.

Mathew 4:10 This is a smart token able to pull through business function such as you go to Apple Wallet, you open up your car token, and you push the button to start your car from the token. You push the button to lock the car from the token. You’ve integrated the insurance token into that token. Let’s say that you have to go, you want to go drive for Uber.

Mathew 4:29 You go to Uber’s registration. They say prove that you have a car of a certain type and that you have certain insurance. To do that, you simply present that smart token and it allows all those systems to integrate together. Now, let’s say that you are going through that Uber flow and Uber says, cool, you’ve got the right car.

Mathew 4:45 You do have insurance, but you don’t have the right level of insurance. Now, traditionally, that individual would have to then either go to their insurance application or their insurance provider’s website to then top up their insurance to the correct level of insurance for ride share. In a tokenized scenario, in this scenario, that service is pulled through into that moment because when you’ve connected that token to Uber’s site, you now have business services delivered directly to you in that moment.

Mathew 5:13 So you simply click a button, top up your insurance, and now Uber sees that you have the correct level of insurance and you’re allowed to be an Uber driver. So we’ve taken those business services out of those centralized places. We’ve tokenized them and allowed them to become portable and composable, creating an entirely new user experience for the web.

Why not build the service with centralized APIs?

Watch this part · 5:31

Kevin 5:31 Love that. Thanks for the example. When I think about the car example, right? Alfa Romeo, if I’m not mistaken, has already integrated an NFT token, basically a non-fungible token, into their car for more kind of traceability use cases, right? Well, the car owner did that maintenance as you know suggested by the manufacturer and so on.

Kevin 5:58 And I really believe that these kind of use cases, especially when it comes to, well, proof that you have that insurance and these kind of things, absolutely makes sense. You also talked about, and I think that’s the kind of special stance on your side, when it comes to executing, let’s say, actions on top of the NFT or token, whatever standard it follows.

Kevin 6:31 It feels a little bit like I can’t imagine a lot of people now jumping up and thinking, well, why can’t we do that in a centralized way and why is that better?

Mathew 6:47 Yeah, that’s a great question. So let’s just break that apart really quickly. So the internet currently functions and we have APIs, right? We can build integrations and we can do these things. Building an integration on a one-to-one scenario is not complicated. If in that scenario, if Uber only had to integrate to one insurance company, that’s not that big of a problem.

Mathew 7:07 They could do that. How many insurance companies are there? Right now you see the problem, which is any time we try to do integrations at scale, they’re hampered by the ability for us to literally scale those integrations. We would have to build APIs to every insurance company and then manage all those APIs to every insurance company.

Mathew 7:25 If we instead move that to tokenization, we can remove that issue of scale. So this is the concept of it’s very easy to do these things one-on-one in a Web2 environment through traditional technology. Anytime that we want to start to scale these partnerships, that’s trick number one. It’s very hard to scale.

Mathew 7:43 Trick number two is data. So anytime that I want to then pass data to you, there’s this thing called GDPR, which means I have to comply to GDPR. What we’re doing with tokenization is actually putting the data in the individual’s hands and the individual is becoming the integration point, which is completely GDPR compliant.

Mathew 8:02 So the individual is the one bringing their data around. It’s not one company sharing via an API with the other, which has data privacy issues. It’s empowering the individual with their own data so that they can take it to wherever they need it to be fired and function to then create better end user experiences.

Mathew 8:19 Those are the two things that you can’t do in a Web2 world.

Must tokenized data live on a public blockchain?

Watch this part · 8:22

Kevin 8:23 The second part is particularly interesting and I’m curious how you view that, on how we can actually do that, right? Because I assume now that you want to have this tokenization on a public blockchain, right? Which means Ethereum, whatever, or any kind of Layer 2, Polygon, or some other sidechain. And when we have public blockchains, data is actually public, right?

Kevin 8:50 So, right?

Mathew 8:50 Correct.

Kevin 8:51 Do you think about integrating some kind of zero-knowledge technology? Or what’s your approach or your thought model in that regard when trying to achieve GDPR compliance or better data privacy, for example?

Mathew 9:08 Yeah, it’s a great question. There’s two easy ways to think about this. Number one is not everything has to live on-chain. Just because it’s Web3 and it’s tokenized does not mean it has to live on a public blockchain, right? There are decentralized ledgers that are not public, right? It still can allow for these things to happen.

Mathew 9:23 That also allows for us to create frictionless experiences without wallets, right? So these tokens don’t require anybody to even have a wallet. We can put these tokens directly in an email account. I can put these directly into an Apple Wallet, right? So let’s first off remove the notion that everything in Web3 has to be on chain in a public blockchain because it doesn’t.

Mathew 9:43 The second aspect is then to the question, so one is the public blockchain, the second is the individual is the data center, right? So they are bringing their own data, which then allows that to be compliant. So that’s really kind of the answer is that it doesn’t all have to be on a public blockchain.

Mathew 9:58 The third is then, yes, ZK. And that’s like a [unclear] statement, but it’s even further than ZK, right? We’re thinking about attestations. So this is another type of technology. An attestation is just a proof that you have something. You can think of it as the same function, which is if I go to you in a traditional world, what most people know is I have a token and I hold that token in a wallet.

Mathew 10:20 Then I go to use that token. I then connect my wallet to a site. That site reads my wallet and then sees what tokens I have. That is one scenario. That is not the only scenario that is possible. You can also have an attestation, which says that I go to a site. I then give a token that proves I have other tokens, which doesn’t allow you to see the actual token.

Mathew 10:43 It’s just a yes or no. It’s a binary, just like a ZK, right? And so we can use those as well to create and share data, so it’s all GDPR compliant. But the easy answer is not everything has to be on a public blockchain.

Kevin 10:53 Yeah, that’s where ZK and other types of cryptography, of other type of mathematics, for example, can really help here, right? For example, so going back to your first argument of data interoperability, I’m curious how you view, because I know people that, I know what some people would say here now, is, well, you still have to agree on a blockchain, on a network, what kind of token standards to use, and so on and so on.

Why standardize a token rather than a REST API?

Watch this part · 10:54

Kevin 11:34 So you still need to agree on some kind of standards. And if we agree on decentralized or centralized standards, actually doesn’t make much of a difference, just that, well, the other thing requires Web3 developers and the other thing Web2 developers, and one thing is centralized and one thing is decentralized.

Kevin 11:54 So in terms of the data interoperability aspect, I would be curious on how would you argue to someone who would come up with an argument like that, for example.

Mathew 12:08 I’m not even sure if I really follow that argument. Can you make it a little bit more clear?

Kevin 12:12 Yeah, sure. Sorry. Well, in general, when you say it’s about, you have, for example, a token, right? And that’s a unified interface. Every insurance company follows the same standard, right? And car companies, for example, or whoever, can then, in a standardized way, can grab, access that data, and utilize it.

Kevin 12:39 And the thing is that you still need to agree on what kind of token standard to use, what kind of network to use, and then people could argue, well, why not agreeing on some kind of REST API interface or something like that?

Mathew 12:57 Got it. Yeah. So good question. So a couple things here. One is, company Smart Token Labs I’ve been advising has created Smart Layer to do just this, right? It is an interface and that allows for tokens to then be interoperable. You can essentially interoperate on-chain, off-chain and cross-chain without having to then build those bridges to the different chains.

Mathew 13:15 So it is an interoperable standard. Now, there’s two other things with that. Part of that is they’ve actually created a standard called TokenScript, which would actually be the interface in language standard. And they’ve also put forth multiple EIPs, the last being, I believe, EIP-6159, which essentially allows for a standard for a smart contract to actually be able to reference and carry through an external script.

Kevin 13:39 So similar to the xNFTs that I heard a couple of months ago, not sure if you know them, it’s not on Ethereum, if I’m not mistaken. I think it was a Solana type of standard, but it was kind of in the same direction, right? You had an NFT that had some executable code in it. So that’s definitely a direction that I [unclear] imagine that a lot of people are targeting at.

Enterprise adoption beyond assets and loyalty points

Watch this part · 14:09

Mathew 14:09 Yeah, I mean, we have to. If we want to reach mass scale, here’s what we have to do. Currency is cool, but there’s too many people that are afraid of it. I used to work with every major Fortune 500 company on their Web3 projects and also Salesforce running the studio. They all want to do Web3, and the CEOs believe in Web3 and they believe in blockchain.

Mathew 14:33 What they don’t believe in and what they’re scared of is they’re scared of asset class tokens because that then creates lots of issues for them. It creates legal issues. It creates regulatory issues. Consumers don’t want them. There’s been massive backlash. There’s brand reputation risks. So if we imagine that these businesses are going to use tokenization in that format, they’re not.

Mathew 14:53 In fact, a lot of them have pulled their projects back. Look how many have scaled down NFT projects in the recent months, right? That’s just not what they’re doing. Now, they’re all still holding on and saying, we believe that this technology is valid. And what they want to use that technology for is solving existing business problems.

Mathew 15:11 Tokenization as an asset doesn’t solve an existing business problem. It creates problems to solve. It doesn’t solve existing ones. What we’re doing here is we’re saying, listen, you have existing problems of how do we integrate and how do we create a better business network, right? Think about network effects and efficiencies, business efficiencies.

Mathew 15:27 Anytime a technology allows anybody to do something quicker, faster, easier, it always wins regardless. What we’re doing here is enabling business networks to share data and create better business experiences and better business efficiencies. Those are two core problems that they all have that they’re trying to solve.

Mathew 15:43 So that’s why everyone is going to be starting to focus on tokens as ways to solve existing business problems, because that’s how we get to mainstream adoption. You take that one step further and you say, once we can do that, well, then we don’t even have to use the same concepts that we’ve used before.

Mathew 16:00 This concept of you have to have a wallet to start. We can use off-chain tokens. In fact, we’ve just done multiple projects where we used off-chain tokens. The individual in the future, we’re not going to be saying Web3, right? Every business is going to be using all types of technology. One, two, three, they still send letters, we don’t call them a Web0 company, right?

Mathew 16:21 So just because you use the blockchain, it’s not going to make you a Web3 company. It’s simply going to make you more efficient and allow you to compete and have better competition in the market. That’s what it’s going to allow for, right? So I think if we keep those things in mind, I mean, that’s where we need to go.

Mathew 16:34 And that’s why everyone’s going to be doing those because that’s where the problems exist. So if we go back to all those CEOs and say, hey, listen, this is not Web3 as an asset class. This is not Web3 as tokenized loyalty points. They don’t give a shit about that. What they care about is how do I create better end user experiences and how do I create better business efficiencies?

Mathew 16:56 That is what we’re solving here. And that’s why everyone’s gonna be focused on that aspect of tokenization moving forward.

Kevin 17:00 I love it. We have mentioned some kind of identity use cases a couple of times, if I’m not mistaken, right?

Can public metadata still reveal private information?

Watch this part · 17:02

17:13 Yeah.

Kevin 17:13 And we also scratched on top of data privacy. So one thing that already happens, and I would be curious on your opinion on that, is even though that a lot of data might be off-chain for all kinds of identity tokens, you still have some kind of metadata attached to it, such as who issued the token, what type of token is it?

Kevin 17:44 Maybe it has a name and some other data attached to it that’s publicly available. And one thing that will be possible and already is possible, that you can tag a lot of addresses and have some kind of, well, you just have a data [unclear] that’s public, right? Even though a lot of sensitive data is off-chain, you might still be able to assume certain things, thanks to data science.

Kevin 18:18 How do you, because it sounds like you’re actively working on that, you’re heavily involved in it, have you already thought about that, or how are you guys solving that, or is that something that’s further in the future, or it’s not an issue at all? Would love to know.

Mathew 18:37 I mean, it’s a really good question. It depends on, there’s lots of different things here. Number one is, depending on where the person, whoever creates the token, depending on how they want to store the data and where they want to put the data, it’s going to be up to them, right? So if they want to put data on public, you know, stuff on a public chain, they’re going to have lots of issues.

Mathew 18:55 You should never put information like that on chain that’s public, right? That’s why we’re talking about off-chain data and those databases. We’re using cryptography and Web3 technology to verify, but you’re only going to be able to verify if you’re asking questions, and then you’re still going to have a, you know, a layer where you’re [unclear] actually going to be able to take the data, right?

Mathew 19:13 So scenario, you take token to company Y to prove that you have X and to pull in business function. They’re going to have a very limited view. They’re just going to fire things. They’re not going to be able to see the underlying data and pull that into their database and keep it, right? So we’re talking about layers of abstraction.

Mathew 19:30 The keys here are still to ensure that the individual owns the data at the end, right? If all we end up creating is a way for somebody to go in and create a bot and use data science and AI to scrape the web and to create user profiles of individuals they don’t want created, we’ve not done anything good.

Mathew 19:47 And so what we’re doing here is we’re really focused on making sure that when we do create these things, that we are solving for those key issues of data ownership. And to us, data ownership just means that the ability for the individual to have the data where they want it and to use it how they want to have it happen.

Mathew 20:02 So that’s step one. Step two is to make sure that there’s privacy in place, which that’s why we talk about off-chain tokens in combination with on-chain tokens to then prevent some of these use cases where you can just scrape everything and then build user profiles. And then the last is then making sure that when we implement these solutions, that we understand how we’re implementing them and we’re not requiring people to do things such as connect wallet.

Mathew 20:23 We don’t need to connect a wallet to a site to know what’s inside of that wallet and to use that wallet. There’s other technology that can allow the individual to stay anonymous from a wallet ID perspective while still validate what’s in that wallet and still get the functionality they need from those tokens in that experience.

Mathew 20:39 So I think it’s a combination of making sure we know what’s possible, making sure businesses know when they create these things, what is safe and how do we create these. And then three, making sure that we build systems that when consumers want to engage and leverage these tokens, that we’re building systems that allow them to then maintain their privacy.

Kevin 20:57 What do you think, what kind of companies, except for sure Smart Token Labs, might get, let’s say, get that right really soon?

Project examples and privacy claims

Watch this part · 20:58

Mathew 21:10 Yeah, I think it’s not going to be that hard because, you know, people aren’t just going to just go at this blindly. No company goes in and says, I’m just going to make a tokenization project that doesn’t talk to people. And so it’s, you know, making sure that as we guide these companies through, that we’re ensuring these things to happen.

Mathew 21:28 And we are. We’re releasing projects with, you know, one of the world’s top cricket players. We’ve done projects for FIFA World Cup. We’ve done projects for, you know, the Ethereum Foundation. There’s lots of big players who are creating these, wanting to do these technologies and implementing these systems.

Mathew 21:43 And they’re abiding by those rules. So, you know, we’re seeing privacy be maintained. And that’s one of the big aspects of the technology that we have in the security is privacy.

Three takeaways on tokens and composable services

Watch this part · 21:53

Kevin 21:53 I love it. Well, from my end, not sure if you have any things that you want to share at the end of the podcast now. Otherwise, I would say to maybe have three key takeaways for people to remember, because we have discussed a lot of different but still closely related topics today. And I think there’s a lot of essence to it that people might want to remember after this.

Mathew 22:18 Sure. Yeah, the three key things that you should always keep in mind moving forward. Number one is that there’s a difference between securitization and tokenization. Right. So and that really means that there’s a difference between tokens that they don’t always have to be assets. They can go beyond just being assets to become applications.

Mathew 22:35 That’s step one. Step two is as we think about the future of the internet, we must think about composable services, right? And that’s what we can then use tokens to do is be composable services. And why would we want to do that? Well, it’s because we’re solving the key problems that the internet has of integration and data sharing.

Mathew 22:51 That’s what allows us to solve those things. And then three is to just imagine a new world where we’re not essentially just using tokens as currencies, but we’re really using them as applications and creating better user experiences. And that’s really the end game here is the businesses that learn how to use tokenization to create better user experiences are going to have the same market effects and dominance that the same ones that learned how to use digitization first.

Mathew 23:14 Netflix, Salesforce, Amazon, right? They were the first ones to deliver services via the web. Then you got to the app evolution, and that’s when we see Instagram and all these app-first businesses. The next wave of businesses hasn’t been created that’s going to dominate the next digital landscape, but I guarantee you they will be tokenized.

Closing remarks

Watch this part · 23:33

Kevin 23:33 I think that’s a great ending. Nothing on my side to add. Mathew, really thank you for sharing all that. If you want me to put some kind of links, I will add them in the video description as well, for everyone watching or listening to this. And I would say thank you, Mathew, for your time. And yeah, thank you for watching.

Mathew 23:58 Great. Thanks for having me, Kevin.

Kevin 24:00 Sure thing.

Jingle 24:01 Wavect, the Web3 software company that understands what you want.

Transcribed from the original audio with Whisper large-v3 and AI-reconciled with the original automatic captions. Unclear speech is bracketed; speaker labels and translations are editorial aids. No complete human listening pass. The recording remains authoritative; statements reflect its date. Published in August 2023. Product, project and adoption claims, including those about Smart Token Labs, Smart Layer, enterprise Web3 programs and the named projects, reflect that time and are attributed to the speaker. Sweezey cites EIP-6159 with the hedge "I believe"; the client-script proposal he describes matches ERC-5169. Claims of complete GDPR compliance or wallet anonymity are not established by this interview. An attestation is not automatically a zero-knowledge proof, and tokenization does not eliminate APIs, shared semantics or integration work. The car, Uber and insurance flow is an illustrative scenario, not a verified deployed integration.

Questions this episode answers

He means a token that presents usable functions, beyond an ownership record or tradable asset. At 03:58, his car example combines access to vehicle and insurance services. It is a proposed interaction model. Delivering it requires supporting software, service providers and access controls; ownership alone does not make the services work.
No automatic replacement follows from the interview. At 12:12, Kevin asks why businesses cannot agree on a REST interface. Sweezey points to Smart Layer and TokenScript. That shifts how functions are presented to users, while integrations, shared data meanings and provider cooperation still need to be implemented.
TokenScript is a framework for presenting and using token functions through compatible clients. ERC-5169 is a contract interface that locates associated client scripts through scriptURI. At 13:24, Sweezey says, hedging with “I believe”, EIP-6159, but the described mechanism matches ERC-5169. Linking a script and running its interface are different responsibilities.
No. At 09:08, Sweezey describes off-chain data, nonpublic ledgers and email or wallet-pass experiences. TokenScript can include off-chain functions. A client-side interface is different from blockchain execution; the implementation must still define issuance, authentication and access to the underlying services.
No. At 10:04, Sweezey describes an attestation that proves a fact without showing the underlying token. That is the disclosure goal of his example, not a property of every attestation. A yes/no response alone does not establish zero knowledge, unlinkability or protection against metadata correlation.
No. At 07:44 and 09:43, Sweezey makes broad compliance claims. Keeping information off-chain can change the design, but it does not settle the lawfulness or privacy of the whole processing activity. The public-metadata question at 17:53 remains relevant. Assess the actual data flows and parties rather than treating token ownership as certification.

Want this kind of thinking applied to your product?

We build MVPs and act as fractional CTOs for founders who’d rather ship than talk.