Why Every Supply Chain Blockchain Died: Capture, Neutrality, and Lessons from Inside

Transcript

[music] Hello everyone. Thank you very much for coming. I'm going to look at something that is a rumor for 10 years now or so that blockchains are really great to track help tracking supply chains. But for some reason that seems to have never worked out.

It's not a idle question. Um there was great hope I mean there was great hope at Defcon Zero actually that supply chain could I I was really hopeful supply chain could be the killer app for Ethereum. turned out it was selling tokens. But at that point um when it was not quite clear, we knew already what supply chain would be good for. Tracking supply chain could help to free people who are at this point in slavery. And it's not a few people. It's around 30 million people where even though we have legislation now in the European Union

and other countries are working practically in slavery. Then we have 1 million more than 1 million people who are dying from food poisoning cuz they are eating stuff where nobody really knows where that came from. And that has consequences that are going beyond that which I'm also going to talk about. Just in the US for example, you have multiple events every year where a farmer gets their entire crop destroyed because somebody ate something in a restaurant, got sick, maybe even died, and then the agencies start to try to find out where

that food came from. And very often it can only be tracked round about and sometimes it's even the wrong crop. Sometimes people think it's tomato why somebody got sick and later turns out actually it was cucumbers. But at that point a lot of farmers have also already gotten the order to destroy their entire harvest. So having something that would actually be able to track in a better way than what we have today would be really valuable and blockchain was built for that. And that's not coincidential because facitiously you could say Bitcoin is actually a machine to track

tokens. And the way that Bitcoin manages to track coins is something that you could just use for tracking anything else. You just already have the ideal machinery built into a blockchain to track stuff because that is how Bitcoin implements the double spending prevention. But still, I mean, even though that seemed to be very obvious to a lot of people and a lot of really serious attempts have been made, um it never quite worked out. And I have been part of two very serious attempts of bricks and mortar companies initiatives coming from the real world so to say where the one was about

tracing diamonds to make sure that people don't buy blood diamonds which means like diamonds that were sold to finance civil wars. And the other was basically that I was the original architect of IBM's Hyperledger Fabric that was also a blockchain design for business to create um a blockchain that is basically something that is specifically for industrial application.

So what was the problem? basically every time and not just in these projects that I have u been part of the party that was most interested in blockchain was actually the middleman who would have been disintermediated by the blockchain and unsurprisingly invariably they found oh well blockchain actually doesn't really help us [laughter] and so the projects died I'm I'm I'm simplifying that a little bit but this is really a pattern like those who were most interested in finding out what this blockchain thing really is were often those who were who felt most threatened by it. And that is

of course a pattern um to make sure that basically the initiatives are set up to fail even if they are actually serious like what I've been taking part in were serious projects but they still ran into the problems that were basically uh mistakes that were baked in in the very beginning. A similar project um so to say to uh something like Tracer which was the diamond blockchain might have been started by a neutral party and not by the biggest diamond producer in the world. In this case it wasn't the middleman but it was just the

biggest competitor that created a blockchain and invited everybody on board. That also kind of didn't fly even though the official reasons are different. But you do have a problem if you have like 80% of the market and then invite everybody else to join your blockchain. And the problem is not the blockchain technology. Obviously, it's just capture by any other name because in the end, if you propose that maybe everybody gets as many votes and as many tokens as their market share, [snorts and laughter] well, nobody's going to join your blockchain.

Now another thing I've been part of as I said um was Hyperledger and Kora. Um Ka I saw from afar so to say. Hyperledger um was something I have been actively designing as architect was the idea of not building in a token. So you have a blockchain that doesn't have a token. Of course, that's an interesting proposition and of course that's something that banks for example really like and if you think about it IBM's main customers are banks. IBM lives from those huge contracts for mainframes from banks still. It's only logical that IBM would come with a

blockchain that doesn't have a token [laughter] because that could be something banks don't really like. And there were a lot of ideas about how a blockchain even without tokens can be very super valuable. But in the end uh I hope I will be able to show that you do need the atomic transactions that you can do that are basically the core of the token transaction to be able to have a viable business model. Why you should use blockchain in the first place? Because otherwise you really just have an expensive database where at some point somebody is going to pull the plug and say, "Well, that actually isn't worth it." And they're right.

So in the end um there's also this question of how much does it cost to have a business model with a blockchain in it and I like to say when I was doing consulting well yeah a blockchain is like a million times slower than if you do it with Java and use a database [laughter] which is probably true and this kind of equation you have to think about and if you are proposing something that everybody should take part in you really have to use everything that this technology offers so that in the end you have a value proposition that everybody can take part in. You cannot just focus on a certain thing that is kind of almost just a nice to have because you are dominating the

market at any rate or you are the middleman for everybody at any rate and then hope that everybody comes on board even though it costs 10 times more than what they're using at present. So it's pretty clear the graphic on uh the left hand on the right hand side is basically showing you it has to be a collective effort. Blockchains are just more expensive if you have the trust of everybody at any rate just use a database. But if you can collect everybody in an industry behind you then you stand a chance to actually really do something using blockchain technology.

Now why didn't that happen? Although it's so obvious. Well, a big part of it is that it's really just not the thinking of uh the participants of a market that is highly competitive. But if you if you have this quality basically that you need to exchange data or you need to agree on data that everybody knows is correct is the truth or is at least that what somebody has at some point signed off on. You always run into this problem that either it's too transparent and then nobody will put their data on the blockchain. the companies don't are not

even allowed to do this because that would go violate monopoly um legislation or you have secrets that nobody can verify. So you have to have something in the middle where you have data that is not out there just wholesale for everybody to see but also not encrypted wholesale so that nobody can verify it. And this is why I'm proposing that there has to be some thinking that is absolutely normal in the normal database world where access to individual data points down to the individual fields like this column and this row is in very

laborious work defined down to certain roles who can see what in a giant database. This is not something that is app developers usually don't need that like if you have a local database you have very restricted rights um very restricted uh very safe uh environment where you don't necessarily have to define rights on the level of the database right that's just not what happens but if you have a big ERP like what is really happening in in the bricks and mortar world with the real big databases that are the backbone of businesses then this is just normal You define roles and you define access rights and it's a really complicated

mesh of things that is then in the end the definition of who can see what when. >> Something that we know in the database in the in the blockchain world is that a proof of absence is actually pretty important or at least this element of absence. We know this since Bitcoin because you can basically only spend a bitcoin when you have a transaction that hasn't been spent by anybody else. So using Merkel trees this is possible and this is something that is very valuable in supply chain that has been heavily underused but this has to be abstracted

in a way that everybody can use it and know that when there's a query on the level of the data what comes back if it is as zero if it's nothing that this really means there is nothing there and this is a level where I have been always warning uh in consulting jobs, okay, this is what happens if you think about it blockchain like a database. You expect all these nicities to have been there, but they are usually not there because for a single project, it is just too involved to create all these things.

And this is why just like with this definition that exists in ERP databases down to the single field. We have to have a view we have to take this hard problem of who can change what who can read what. How are we going to make sure that a single cell can be read by different people? wrapping it in different keys um is something that we are having at the lowest level basically so that a tool exists that is abstracting all this away and we can think freely so to say on the blockchain level of what the application actually can be and one important part that is my

learning um about what works and what doesn't work um using a blockchain for real life real businesses is you have to use all of blockchain. I said this before it is really important that you don't reduce it to a database use the transactions that you use everything that a blockchain can offer if you want to have a value proposition that works for a supply chain. And if I just come back to this example with um farming in the United States, the reality is that farmers ideally want to have sold their harvest before they even put the grain into the ground. That is just normal.

That is just part of the business. That's where the word future comes from. That's where future exchanges come from. So obviously blockchain is the ideal instrument to have trustless contracts that are involving this type of pre-sale of the harvest. Why would you leave that out of it? Why would you leave that out of the picture when you try to create a blockchain that is depicting the journey of perishable food from farm to fork? Finally, that's another uh passion of mine, the question, what

is a smart contract? So, I'm proposing that we will see smart contracts that are written in an in a language that is actually reading like normal English. This is possible today. Um, there is a language out there, it's called Lexon that I created that has this capacity that a smart contract can look like this. And this is something that a farmer and a restaurant owner can sign and a judge will accept that this is actually a contract. There are some jurisdictions in the US where a judge will say, "Yeah,

okay. A contract is can be on the blockchain, can be a smart contract, everything. But a contract usually is the meeting of the minds that a judge finds that two parties at some point in time wanted the same thing. And the paper is only the proof for it. But I believe that smart contracts in the future will be written like this. And this is a huge part to make supply chains work because on the lowest level as I said futures selling your harvest before you even start working are completely normal. It's not something futuristic for farming.

So this is how a contract looks. A contract a grower can read and most importantly also a lawyer and also a judge. And that makes it something that is tangibly existent in the realm in the legal realm so to say. Now this slicing and dicing so to say of the rights to the individual data fields is reflected also in a slicing and dicing of capacities of backends. And I'm the protocol lead of Swarm, so I'm biased, but I'm sure that actually the tool

that's out there that is useful for the back end for this type of supply chain, there's only one, and that has to do with how it organically grew, so to say. This has to do with where it actually is at this point. But we do have an answer to this question. Where do we actually store the data? And it's swarm. There are reasons for that. We swarm had a great presence here at the booth. Um I can I recommend you to go over there and look at what swarm can do now. It's really grown beyond being a hard disk and a passive storage. But there are for supply chain that's not even that interesting. It's the very simple model how you have price building and swarm.

Why that is the tool of choice if you want to build a supply chain and want to have the data in a trustless accessible way stored online not just on your local computer. So to recap real quickly, the principles that I found that you have to apply to create a successful supply chain project finally on blockchain.

First of all, nobody who trades on it should actually run it. Second, you have to use all of blockchain. And he used this key to success of NFTTS that was basically the transfer of the right to the art and the cut that went to the artists was one atomic transaction. That is really the one thing where smart contracts make the difference. Like nobody could sell the stuff without the artist getting a cut. You have to use this principle for supply chain.

Then everybody has to use it. It has to be an effort for the entire market. It can just not be just the richest participant, the most successful participator. It has to be something where everybody's invited. It has to be open. Finally, you have to go and go down to the level of the fields and the cells like it is done normal databases when you are designing the data architecture that what you actually need on the level of the supply chain. cannot be done with just the tools that we have just saying, "Yeah, well that's what works on the blockchain." We have to go one step further.

Finally, smart contracts have to be readable. They have to become something that exists in the legal world so that a judge or a lawyer parties know they are bound by it. It's not just a technical thing that transfers tokens. We should be using storage that is on the same level like the blockchain itself that the smart contracts are running on. It has to be a distributed decentralized data store. And then finally, it should not be the intermediary themselves that is running these projects who will be disintermediated by it.

So is this going to hold up? I'm going to find out. I'm running a project that's called source link that is doing exactly that the farmtofork element with a lot of bricks and mortar experience behind it following these principles and uh I'm happy to answer you questions at this point what uh [snorts] >> thank you very much >> James do we have any questions Thank you for that. Mic on. Great. You can tell it's late in the day because that made me hungry. [laughter] Thank you so much. Uh we've got a few questions we'll run through. So first

one that's came in was could this initiative succeed in case in a case where the blockchain is built by a market participant not by the not by a market participant but by a regulator. >> Yes. thought [laughter] uh I think a a regulator is a much better is much better positioned to um offer a supply chain or any it doesn't matter in what industry actually uh there are examples actually the European Union has done something very interesting with data about publicly traded companies it was great project they really understood exactly why they wanted to use a blockchain for it because the European Union had in initiated a initi a law actually that

there should be one central entity where you could get information about any publicly traded company in the European Union just it never happened because the countries of the European Union could never decide who would run that server. So it was a great idea that they started this uh blockchain project where they said okay so everybody on a blockchain node we put it on the blockchain and I think that didn't in the it was programmed it was designed and programmed but I think in the end it also didn't happen because for other reasons and that's the butt right regulators can be very slow just because something works doesn't really mean that it politically it's going to happen in

the end so they come with a host of their own problems that they add into the Thank you very much. Um, and then a more general question, but I'm interested in the answer. To your mind, what's the most current interesting user case of decentralized storage [laughter] we to my big surprise um um I found that at Swarm and that's that's obviously what I'm going to talk about. Um, as a protocol leader of Swarm, the most interesting stuff we're doing right now for decentralized storage is streaming, live streaming.

And there are reasons for that. It turns out that the architecture of a network for decentralized streaming is also great for live streaming. Of course, you didn't have it all stored there. And it turns out that the music business is in dire need of getting their intermediaries disinated. So, surprisingly, that's my answer to that. >> Great, great answer. Thank you. [snorts] And next, what's the current state of Lexon and can you tell the audience more about it, please?

>> Oh, yeah, please. Um, lexon.org is the website. Uh, it's not entirely up to date. I have to admit there was a lot to prepare for Defcon. That was uh more to the forefront, but I'm happy to um for anybody to grab me and tell tell them more about Lexon. There's also two books out there. I had two talks. I'm not sure if they are recorded here at Defcon about it and um yeah, it's definitely a thing. >> Perfect. And then uh can you give a short list of supply chains that you think we should be building for >> everything?

>> Honestly, if you think about it that we don't know what is in our food and we really don't know that. >> Yeah. >> It's just incredible. It's just a very much of it is food powder where that it's made for because that's easier to transport and there are initiatives that are b basically using spectroscopic uh technology and so on to try to find out where stuff actually comes from but it's almost unknowable. So this is this is a real bad thing and honestly it's I I don't know exactly how it is in India. I know that it is the worse the more industrialized the food industry has become and it is it is

something that is killing people. So that's really important but there's all types of others where where for example um fishing >> you have the problem that there are slaves being held on those trwers. So it's really important that we get to a point where it's clear where your tuna comes from. >> Yeah. So that this slavery on the seas because it's very easy right people cannot run away uh can be shut down. >> Perfect. And one last one sorry we haven't got it everything but when do you expect the first commercial blockchain for supply chains will be deployed.

>> So I don't think you have to deploy a new blockchain like you don't have to deploy a new storage. It's on the application level what I'm talking about here. And the questions I hope will have become clear. The question why it doesn't work is something that doesn't necessarily has to do with the technology. At some points I do think we have to go deeper. We have to become more user friendly. That's a refrain that we're telling ourselves all the time. But it is also about who actually does it. And I think that is the bigger thing about it.

>> Thank you so much. And just one final point of housekeeping. programming here tomorrow starts at 10:00 a.m. Thank you all for your attention and thank you Henning so much for us.