In this episode of Blockchain Innovators, Conor Svensson - founder and CEO of Web3 Labs, talks to John Wolpert, ConsenSys Group Executive Enterprise Mainnet Products and Technology.
John is also the co founder and steering committee chair of the Baseline Protocol and prior to his work at ConsenSys he was a co founder of IBM's Blockchain Organisation.
John discusses the history of IBM's blockchain division and why he believes the future of blockchain in public. He also talks about the Baseline Protocol, a technology that uses the Ethereum public mainnet as a common frame of reference among different organisations.
John has an acute understanding of what enterprises need from blockchain and he provides lots of insightful take aways in this podcast.
You can also watch this video on our YouTube channel here.
Hi, it's Conor Svensson here, founder and CEO of Web3 Labs this is a conversation I had with John Wolpert. John Wolpert is the ConsenSys Group Executive for Enterprise Mainnet, delivering products, services and leadership to enterprises that use the public Ethereum mainnet to conduct their business. He's the co-founder and steering committee chair of the Baseline Protocol, an initiative by the Open Standards body Oasis and prior to his work with ConsenSys he was an enterprise executive and entrepreneur and was one of the co-founders of IBM's blockchain organization. In this episode we discussed the history of IBM's blockchain division and why he believes that the future of blockchain is public. We also cover the Baseline Protocol. Baseline Protocol is a technology that uses the Ethereum public mainnet as a common frame of reference among different organizations. John really understands what enterprises need from blockchain so I'm sure you'll have lots of takeaways from this insightful conversation. Hey John, it's great to have you here. It's good to see you Conor. So what I wanted to start with, origin stories are I think, always a great thing when it comes to people talking a bit more about themselves and what's brought them to where they are. But yourself, you have had a significant career at IBM before you moved into the direction you're focusing on now. I'd love to unpack your journey there, especially what led you to being one of the co-founders for IBM's blockchain initiatives and then how you ultimately moved on from that to work on what you're working on right now. Well, I guess I was an 'IBMer' three times, starting in 1997 right out of grad school. I worked on the Java team as well, you know the original Java team out in Cupertino and making Java safe for business working with Sun and and others on that and then I left to start a startup and then came back to help start Extreme Blue which was a big incubator kind of thing that we did and then left again, moved to Australia to work on something that IBM participated in. It was a sort of an NGO, non-profit government-backed organization to help do open innovation between legally separate entities where research departments and different universities and different companies could collaborate without exposing internal secrets to each other. It was kind of a fun gig and then started Flywheel, which is one of the original ride-sharing companies. This was focused on taxis right around the same time that Uber was coming up and got our butts kicked by that and then came back to IBM again and went to work for Jerry Cuomo, my favorite boss ever. Yeah, one day he called up and said 'Hey Louis, it's Clark, there's this thing called blockchain we're supposed to go check it out!' and I tell the story a lot. You know, I said okay give me a shipping crew and I'll go check it out and he said 'Well, I got a musket, two dogs and a canoe!' And yeah, we just got lucky. It turned out to be a something that the chairman and our CEO, Jane Rometty, at the time and also our the current CEO, Arvind Krishna, were both directly involved with and they tapped us to go and do the work. At that point, it took a while before Fabric actually came out. Sorry, IBM came out with Fabric as well and they were looking at some of the public Ethereum technologies and to take there right? Yeah, yeah. The first gang of us we all went, I was living in San Francisco at the time, in Berkeley and we were at the Galvanize Incubator Space in... was at Harrison or Howard? One of those in Soma and they gave us like a broom closet or something like that. There were tables in there like tables on racks in this basement room down there and we just had maybe 10 or 11, 12 people there. A couple of really strong Ethereum proponents were in the room and we say, 'yeah, okay let's go try to start this on Ethereum' and we spent the whole summer of 2015 trying to make it do what we wanted to do, work on Ethereum. In those days we kind of had this notion that blockchain was like a fancy database and that was, you know, it took some years for me to kind of come to my senses and realize that was probably a bad idea. Yeah, but that was the journey. To say okay well, we don't like gas, we don't like this, well let's rip that out, let's rip and then I think Ben Nguyen, one of the original senior engineers on the project, I think he really wanted to show what he could do and you know, he was very skilled and so he liked the idea of chain code. Then Gary Singh got involved and Chris Ferris and others. Shortly after that, I think we decided over a lunch with a London stock exchange group guy named Moiz Kohari, really nice guy, and it wasn't a boozy lunch but I think we'd had like one drink or something. And, we were like you know we should start an open source foundation. Yeah, yeah, yeah. And then Chris ferris and his boss convinced us that we should go with the Linux Foundation for that. That was the start of Hyperledger. Yeah, awesome and from there though, obviously IBM created Fabric and they provided some interoperability a while back, but it was kind of very much - target our enterprise customers, it's a classic IBM type product plate. Well, the problem there was, if you don't mind me jumping in, sorry. But, you know the problem was you have to be careful who you select as your sponsor users, that initial five companies that you're going to listen to while you're designing the product or the system. And you know why it's called Fabric? It was called Fabric because I was running around telling everybody, look we're not you know Corda and others Corda was originally on the thing and I said Fabric is called Fabric because it's not a platform, it's supposed to be the Fabric underneath platforms. If you could build Ethereum or Bitcoin and half the time because of Fabric, then we did their job right. That is not what wound up happening, that's not what we wound up doing. It was my original intent, I think and you know and others, but that was the kind of the initial idea and that's why we called it Fabric. The reason it's still called Fabric is because I think, Mark Parzignak, one of the guys on the team who was in charge of the repo, just named it Fabric and it stuck. We didn't go through marketing, there was nothing, it was just called Fabric and that's why it's called Fabric today. But obviously, Fabric has become a platform and because we had customers like the GTCC and others whose requirements looked a lot like a database, they needed a database, they needed compartmentalization, they needed access controls, but yet we wanted to call it a blockchain and we wanted to have distributed multi-party governance and all that stuff. And, as it turns out, I remember the whole time Mark Andreessen was laughing at us, you know over in Silicon Valley, he would make comments about how this private blockchain movement was kind of dumb and you know what, he's smarter than me, he was right. What you get is a complicated shared database. Not that shared databases are a bad idea but, we should have just called them shared databases. Yeah. And the reason you should call them that is because then you realize you are putting data on a very large threat surface you know, even if it's private right? So if your mental journey was - I really like blockchain but the public blockchain is a terrible place for me to put our internal sensitive data, even encrypted and... I don't have access controls and all these other things because blockchains don't like that. You know, a tool has its own intentions so you want to be sure that you have that in mind when you decide to use a tool. In that, in this case, the blockchain is - I like to spread information out, I like to decentralize it well and if you want to distribute information, use a distributed database if you want to decentralize information, then you can use a blockchain. But what we were thinking about was more like distributed database and we had Cloudant already, we had Cassandra and other ways of doing crash fault tolerant distributed databases. Using a shared database can be a very useful thing as long as you understand that either one of two things is happening right? If your mental state was - I don't want to use a blockchain because I want to put data on that public thing that everybody can see or we'll use a private permissioned one and that's good because all the counterparties will be able to say that they have a node right? Well, it turns out that a. that is complicated. They don't really want to run a node, typically the kind of companies that are going to get into this have already accepted the idea of cloud so that's where you see things like Kaleido and other things popping up to say 'hey, here's a blockchain as a service'. Well I say, if you can Blockchain as a Service then you could probably just 'SaaSed' right? And unless you really have a good reason that all the counterparties need to have their own node and run their own node and I found that very rare to be the case, most of the time they were perfectly happy to have IBM run it, in which case we could have written, all we needed to do is call it blockchain and then written it in Mongo. It would have been more secure and more inexpensive for the customer in the end. It wasn't like anybody was trying to fool people out of their money, we were just on a journey. We were trying to figure things out. Yeah. And so, that was what what wound up happening. That was my observation of the whole thing, it felt a little bit like the Emperor's Clothes right? That we have the shared database and what happened, well we had all these people, all these companies that did POCs. Some of them said that they're in production, some of them still say they're in production but, honestly there's still a lot of training wheels and guardrails on those things and so it's not like they're very central system of record type of positions. In most companies, I would say and the reason is that the CSO gets involved right, the chief security officer says 'Wait a minute, What? You want to put our information on what? You want the stupidest administration system admin in our consortium to get hacked and wreck us all?' Right? And that's the problem. Even private blockchain, if it's actually a blockchain and different companies are running their own nodes you have a security issue. You have a giant threat surface so that's why around what 2017, I was like wow, that doesn't work for me. I was having a hard time thinking about it and Joe Lubin was like 'Hey you know, if you want to come work on blockchain come over here'. That's what I did and that's been a really good experience. But then the problem still remains if you think of a blockchain as a database, businesses can't use either a public or a private blockchain, not for very much. I mean there might be some use cases but you're putting almost all internal data is sensitive data and some kind of in one way or another all business logic between you and your counterparties. You don't want your competitors knowing about and decompiling the business logic in there, you don't want that so everybody says 'Oh let's NFT things' and 'Let's smart contract things'. I'm like 'No I don't want to' and you have noisy neighbour problems and custodianship problems and all these other issues if you're actually using it as your back end to your application, your private internal application. Don't do that. So then well okay, what do you use blockchain for? Well it's real public blockchains like Ethereum that are really good at tamper resistance. They're supremely bad at surveillance resistance right? And they're not great at scale and they never will be right? No blockchains don't scale the way a private regular old database would and even then, you don't have a singleton that can handle all the reads and writes of everybody's application everywhere all the time. You hear stuff like coming out of Dfinity, I like Dfinity it's interesting, but they're not a blockchain right? They're a bunch of subnets they're like virtualized heroku. Which is fine, that kind of interests me but, don't call it blockchain. We might have some blockchain sauce in there somewhere between the subnets but it scales because you're dealing with lots of subnets. A singleton does not scale. And you've spoken previously about how state in effect needs to be scaled for blockchains and that's kind of almost like an academic calculation as to when it can kind of reach these points of being able to handle the sort of loads it needs to be to significantly impact the utility of the internet. Yeah, I don't think that it's the world computer. I don't think a blockchain ever will be. I think a blockchain, and ultimately you need one, that serves all the other ones right? You can have lots of different state machines for lots of reasons but ultimately whenever you need machine A to be able to coordinate with machine B, you need a common frame of reference of some kind. You're either manually setting up an interop right? Or you put an integration bus in the middle, or you primary one or the other machine, or one of the end machines. You have to do something to keep time to this before that and to make sure that everybody you know traffic cop right? Yeah. Well now we have this always on tamper resistant can't lock you out of value valid operations or is this being taken over state machine, that's always on and you get to pay as you go that you can use as the traffic cop. And what do you put on that? You put hashes on it, you put proofs so the final analysis for me is and this may sound a little boring, but for IT people it's kind of exciting, is hashes for blockchains - data for databases. My data and my business logic on databases and systems of record and then connect my SAP and my Microsoft Dynamics and my Netsuite and NEO4j and Mongo and all whatever you have right, Excel spreadsheet, Quicken right? Whatever you're using, coordinate them so that the record in my SAP system is verifiably identical to the record in your oracle system because I'm invoicing you and now I need to be sure that my invoice is your bill and that neither of us can say that we didn't get the memo. We need to have some way of being sure of that is on a supremely tamper-resistant, always-on, state machine. That's where Ethereum is a pretty useful thing, it's a bulletin board. And with regards to the Baseline Protocol, we already have a lot of those integrations. If they're not already out there, there's companies generally working on them as you say, with these very well established systems of records, that are very widely used within enterprises. Yeah, yeah and what I like about the Baseline Protocol is that it's not in a cold war with traditional systems. We're saying 'Hey, yeah leave SAP? Yeah SAP good. SAP on Baseline even better!' Because then you can be sure that your SAP system not only tells you what you think you know about your relationships with other people but you can verifiably know that they also know what you know on a record-by-record basis with really good compartmentalization. Now you have access controls. I've sent you an invoice by a secret peer-to-peer, point-to-point messaging, we have a proof over here that looks like nothing, nobody can generate a classifier from AI on that, it's just a piece of nonsense on a public bulletin board. But that little piece of nonsense, if you and I are in on the secret and know where that Merkle tree is, we can say 'Yep, you and I are in synch'. We can do a lot of cool things from there. We can say it's not just a one-step process. I can say 'well I've got a giant workflow' and I can put rules in to say I'm going to change state from here to here to here. This purchase, this RFP generates there, leads to this master service agreement, that leads to this purchase order, it leads ultimately its invoice, this dispute resolution all these things, these inventory changes. You can start to build really interesting workflows that have integrity. But where like if you and I are in workflows step number 25 ,we may not even know that there is a work step number nine or who's in it right? But if you did that on like say Fabric in a channel or on a blockchain, you would at least be able to observe that other people were doing other things and just the observation of the changes to the machine could give you a lot of intel about what's going on outside of your purview. But with Baselining not so much, there's no information. And with the approach that Baseline takes, a key part of this to say it's using blockchain as an integration layer in the form of a place for system for records to be stored securely but these are more cryptographic proof records as opposed to the raw data as such so there's this kind of degree of security that enterprises like. Then, if you look to what's happening within the public, I guess, the pure kind of play protocols that are building on the blockchain where people are, say DeFi to take a case in point, where they're building these ecosystems out in the open and in many instances trying to take on those enterprises which are finding ways to integrate using the Baseline Protocol with these public ledgers and make it fit with their current way of doing business, do you think that that's going to be enough for these large corporates or there's still significant risk of them being disrupted by some of the big innovations that are actually happening with these public projects that are building on top of the public blockchain protocols? I think that where there is a public protocol, I mean first of all, we should say that public blockchains that want to work with enterprises and any business that gets to the level of being serious right? If you're just a couple of folks in a company, you haven't really begun to grok the realization that the data that you're collecting and managing or that you're dealing with, that the privacy matters, that compliance matters you know? If you're a four-person startup and you're dealing with grandma's pension, you've got all the same regulations to deal with that an enterprise does and you better realize that sooner than later, right? So disruption is fine but, if you're a serious business then you have serious issues to deal with and one of them is you just can't be out there in a digital news colony showing all of your internal information and all of your business processes to everybody. A public blockchain implementation is always going to result in that, even if you try to do things like zk snarks. The only truly private blockchain out there, blockchain structure would be one where it says 'somebody gave somebody something. somebody gave somebody something sameness', over and over and over again. Things like zCash would be like that. That is pretty private but it only does one thing over and over and over again. So, there's no morphology to pick up on and generate classifiers on. That's not what you get in these full-featured blockchains. So I don't know, I'm always dubious whenever I hear somebody with a protocol saying 'We do this much TPS'. Right? I'm like why are you trying to get a blockchain to be supremely good at transactions processes per second? That's not what they're good at. Don't use them for that, if you're trying to use them for that, you've got the wrong idea - you think it's a database. Don't use it for that. It's middleware for businesses, it's middleware. Now with the caveat to that is DeFi. In DeFi, there's a lot of people out there, a lot of wealth management's going on that can be done out in the open. You still have to look at - if I'm a hedge fund pretty sure I don't want my competitor hedge fund knowing a lot about my positions right? And so, that's a problem. What are you gonna do? Okay you throw it through a mixer, now you're in trouble with the federal government. So the fact is you have the nice paper trail of where things went. Anybody with a good machine learning system is going to figure some patterns out about that, so you have to be mindful of that. But that said, again for example, in the company that I'm working on, I've got database records that represent invoices. I'm not going to put those database records on the blockchain but I want to sell a big tranche of them on the public Ethereum mainnet as a token to say 'hey you can have a share in all of these invoices', that doesn't give away any information about like Walmart's invoices you know, it's just a big tranche of invoices with a big rating on it that says 'hey you can buy in on this'. So that's a good example of where you have a traditional system of record that's Baseline so that you know that that data is consistent with the buyer, the supplier, the auditor . Everybody's got consistent information. You didn't have to blockchain it for that. You didn't have to put on a big threat surface. You've got basically four databases that have the same piece of information but then you can roll those up into a token on the blockchain and sell it. So, that's a good example of where you can have blockchain and DeFi working hand in hand with traditional systems. Yeah, that's a fantastic example in terms of really calling out where it should be being used. So in terms of what's really exciting you in the space right now, whether it's Baseline Protocol or other initiatives that you're involved in? Well, I think that for me the exciting thing is how boring we're getting, right?! So boring is the new exciting! Baseline is a pretty old pattern, it's just a loose coupling pattern, I've been using that for what? 40-50 years? So that should be acceptable to CIOs and CTOs and CSOs and the like to say 'Hey, you know, you don't have to be afraid of public blockchain. You just need to use it right and most of what you've been taught about it for the last five years is wrong, so let's fix that', and then you'll be like 'Oh, yes of course, this works.' So basically boring is the new exciting and then of course, I'm being able to do things like finance invoices in an efficient and lower risk way right? If I can bring the risk of non payment down on an invoice, I can buy it for less, I can give you a better discount on that invoice. That's it! Make it all as boring as possible! It's definitely an exciting way to look at it! In terms of if people want to keep up with, what you're up to, I know you're very active on Twitter and you've got a good following there. Is that the best best place for people to engage with you and of course, joining the Baseline Protocol public calls as well? Yeah, getting involved with the Baseline Protocol just go to baseline-protocol.org. We've got a Slack, a lot of community uses the Slack channels primarily for lots of reasons. But, we do have a Discord and Telegram link and yeah, Twitter is good and the YouTube videos, we've never missed a week of the Baseline - we call it the Baseline Show is basically our office hours. In fact, we're just about to launch an India chapter of Baseline. You know, in IT when the major Indian companies that are doing solutions get involved, I mean, if you go and look for SAP education it's probably somebody in India teaching that class, so it's very exciting to see folks in India taking leadership on that side of things. Yeah, awesome. There's a fantastic community there and it's just sprung up so quickly as well, it really is phenomenal. Yeah, cool. Good to see you. Yeah, yeah likewise. John, thanks a lot for your time and it's really great to have you on. You too, cheers. you