We had the opportunity to interview Sam Cassatt at TechCrunch Disrupt SF. Sam is the Chief Strategy Officer of ConsenSys…
We had the opportunity to interview Sam Cassatt at TechCrunch Disrupt SF. Sam is the Chief Strategy Officer of ConsenSys and has great involvement throughout the organization. Sam has an interesting background in that he worked on biocomputation in the field of Cognitive Science at John Hopkins and then consulted around the topic for years.
He went on to work at ConsenSys just a few months after it was founded and has been working there ever since.
Editor's note: This interview has been edited for clarity. The full video interview can be found below.
CCN: How's it going, everybody? We're here with Sam Cassat, the chief strategy officer of ConsenSys. How are you doing today?
Sam Cassat: I'm doing great. Hanging out at TechCrunch and a super cool conference.
CCN: Yeah, it's great to have you on. So talk a little bit about ConsenSys' general strategy going into 2018.
Sam Cassat: So ConsenSys has gone through a lot of different iterations over the last few years. We've been around since late 2014, early 2015 and we began sort of when no one knew the word Blockchain at all. People thought we were a little crazy when we walked into boardrooms and said, "We're going to revolutionize your technology infrastructure, your fiscal infrastructure, revolutionize the economy." No one really understood what we were talking about. Fast forward a few years. Obviously, Blockchain is a buzzword. Everyone realizes that there's some application of this to their world.
But when we started we had to really do everything. It was a little bit like 1991 and we knew that the internet or we knew that the web would be something interesting just like we knew that Blockchain would be something interesting, but there was no infrastructure. In 1991 there was no browser or there was no web server, there was certainly no DevOps infrastructure. So we set out to really build an entire ecosystem at once and that necessitated building our business model as being a venture studio.
When we looked at how to build a whole ecosystem, how to build everything at once. You can't just make one company that makes everything. Because for one, I don't know of any successful example of that. It's distracting. It's not the right level of abstraction to think about it. We felt it was too early to make a venture fund because everyone would have had to repeat each other's work. Right? If I had seeded 20 different companies, all of those companies would have had to build Truffle. They would have had to build MetaMask. They would have had to build Infura. They would have had to build all of these infrastructure components that are shared between all the ConsenSys projects everyone was building on Ethereum and etc. So what we did is we decided to make a company that's at the level of abstraction above one company. It's a company that builds companies. I think we've done that fairly successfully by trying to build things like prediction markets, like Gnosis, things like supply chain platforms, games, etc. That motivated all of the infrastructure that we needed to build. So we've ended up with probably about 50 or so portfolio companies that have been in that studio, in the part of the company that builds companies. That's gone fairly well.
On the other hand, as we progressed, as the market has progressed, governments, Fortune 500 companies, Central Banks entities like that. When you go Googling Blockchain, you go Googling Ethereum, you find us pretty quickly and there's a lot of demand for us to help the entities like that implement Blockchain technology into their existing infrastructure. So we have a part of our company called ConsenSys Solutions that is taking all the infrastructure we've built for our own needs for our own applications and helping other companies and institutions build their own applications that utilize Blockchain technology. So that one looks a little bit like, you know, like an Oracle perhaps. In the early days, you couldn't just buy a database off a shelf and plug it in. You can't buy an IBM mainframe off a shelf and plug it in. You need a solution designed around it.
So ConsenSys Solutions takes the Ethereum Blockchain technology, designs a solution based on it and then implements that. So, we think that a combination of making a whole bunch of companies that are really good manifestations of Blockchain technology that we build and support as well as helping all the incumbents in the world realize the use cases of Blockchain at the same time. That's really kind of the golden couple there. And then around that, we have a ton of other things going on. We have a venture fund, we have a social impact arm, we have an educational institution. Actually just today we announced that ConsenSys Academy is launching their first non-developer course on Coursera. It will be available to everyone in the world through that platform. So we have a lot of things going on. Our general strategy is to just develop the whole ecosystem of Ethereum, get it into the hands of as many corporates and as many governments as possible and get the whole world and realizing what they can do.
CCN: So one of the interesting things about the venture development studio model is a key concern over what I call the Amazon effect, which is, Amazon enters healthcare and there's a 10 percent drop in every single healthcare stock. ConsenSys in Blockchain is comparable to Amazon on the public sphere.
Sam Cassat: Thanks.
CCN: If you know anything about Blockchain, you probably know about ConsenSys and you definitely know about Ethereum even if you're just investing in Bitcoin. So is there any concern over discouraging competition because you have so many spokes on that hub?
Sam Cassat: That's an interesting question. I don't think so. I think in the very beginning, definitely not because there was nothing. And we wanted to make it move faster. So I think certainly there's always competition in any market. I don't think the market would have developed as quickly had we not put up things like Infura and Truffle and MetaMask that everyone else uses. I'm not sure I necessarily believe that we scare anyone into not entering the market. I hope not. But even if that does happen, there are a whole bunch of companies that wouldn't exist if it weren't for those tools and the ecosystem building that we've done. So I think net positive we have a good effect on the market.
I hope we don't scare anyone and entrepreneurship, I've heard people describe it as jumping off a cliff and building the airplane on the way down and people like that, I don't think it's scared off people in general. So I hope not.
CCN: Yeah, that's an interesting answer. And speaking about that, what are some interesting companies you've seen come out of the Ethereum ecosystem but not necessarily related to ConsenSys? Or maybe you guys just did an investment?
Sam Cassat: That's an interesting question. I think Loom Network is a really interesting project. So Loom is a way to basically take DPoS which is the technology used in some other Blockchains that Ethereum rejected because it doesn't have the right security characteristics for us and for what we think is a layer one solution. What they did is they took the base Ethereum and they added a delegated proof of stake on top of it in a framework for adding other ConsenSys mechanisms on top of it that then settle into the base, what you think of as normal Ethereum.
So those top layer protocols can go faster, right? If I want thousands of transactions per second because I'm running a game or I'm running a social network or something, I probably don't want to run every single, like, or click on my social network through the base Ethereum layer because it's expensive and too slow. So what I might want to do is build another layer and then if there's a problem in that fast layer, I can settle that problem on the base Ethereum layer. Does that make sense? Loom Network is doing that in providing a framework for doing that. I think they are a really cool company. I also think a lot of the stake channels work that's happening right now. Companies like FunFair Tech, for instance, are doing a lot of really interesting work in that and really advancing the ecosystem. So I think we want to support companies like that that are working to promote the whole ecosystem. And I think those are some really great examples.
CCN: So like, Loom, right there an off chain scaling provider. Is that something that you see is going to be needed 10, 15 years down the line or do you see that moving back on chain as you guys implemented proof of stake and sharding?
Sam Cassat: I think both. I think it's needed right now. Proof of stake and sharding I think will bring us in order of magnitude or two higher than they are right now. You know, if we're at tens of transactions per second, that could bring us to hundreds or thousands of transactions per second, which I think is good enough for some applications. It's certainly good enough for a layer one, especially of a bunch of financial applications.
It's not fast enough to run Facebook on. You're going to need something else that has different security characteristics. At the same time, sending a billion dollars has much different trust needs and much different protocol requirements than a like on Facebook. So I think it's a waste and it's too slow to use the base layer of Ethereum for a like. So I think they're -- just architecturally, it makes sense to make another layer or multiple layers, like in plasma, for instance, that can be much faster and cheaper and more appropriate to the application. I think that's happening right now and we need it right now.
CCN: So let's dig into that a little bit. Do you ever see a like on Facebook being an appropriate use case for on-chain Ethereum?
Sam Cassat: Depends on what you mean by on-chain. On a chain data structure of some sort, maybe. Probably not on layer one Ethereum, though.
CCN: Got it. So that'd be something more for a directed acyclic graph that writes, too.
Sam Cassat: Like a side chain or a DPoS or some kind of other decentralized database that sits alongside the base chain and maybe the root of the application. And the smart contract that is the finances of that social network probably needs to sit on chain, on layer one chain, but the like is probably, it's not as important. It doesn't need billion dollar security. It can be someone else that needs less security, I think.
CCN: So that's another interesting point. As far as distributed database systems go, how do you see that progressing over the next few years? Because as it stands now Ethereum really is not good for a database and it's not meant to be one.
Sam Cassat: Sure. Ethereum is not meant to be storing large pieces of data. It's meant to be storing things that need a lot of trust. It's meant to be, Sam had a million dollars and now John has a million dollars. That kind of thing that we need -- normally we need lawyers and court systems and police to execute in our society and our economy. That's the thing that the Ethereum base layer makes cheaper and better and faster. It is not storing the entire DVD library in my bedroom or whatever. That's a waste of that high trust resource. I wouldn't take a lawyer and have a notary notarize the data of my DVD collection or something. Right? That's a waste of that high trust infrastructure. So, we have other infrastructure like IPFS for instance. There are messaging protocols like Swarm that are coming out.
There are a whole bunch of complementary technologies, many of which are designed to store large amounts of binary data like video data or pictures or large bodies of text or what have you. I think there'll be a plethora of those. We use IPFS for a lot of the work that we do in that domain. And I also think there are other decentralized databases, database technologies coming up that will be useful for different aspects.
CCN: So can you talk through what IPFS is?
Sam Cassat: The interplanetary file system. This is a decentralized basically binary storage mechanism that is content addressable. So what that means is, say I want a movie. I can take the entire movie. I can do what's called a hash it, which is one of the main functionalities that underlies IPFS. I hash it so I can turn that entire movie into a small string.
Sam Cassat: Maybe it's this long on the screen and that is sort of the name of the movie on the network. And it also is the way that I find that movie on the network, right? So if I were to want to find a movie that came out recently, I don't know, Jurassic Park. I want to find Jurassic Park. I would look up in a list somewhere, the hash of Jurassic Park. I would submit that to the network of decentralized protocol. It would go find a whole bunch of copies that are stored all over the world of that particular binary, that particular blob of data. And they would find it and deliver it to me. Right now, volunteers all over the world or people that want to store that data for the rest of the network are doing that. So if I have a business and my business depends upon me hosting data, just like if I have a website and my website depends upon me hosting data, I would have a server and store that data on IPFS.
It's a different way to find that data. So someone could host that. A lot of people would copy that data. So you end up with a decentralized set of copies of that data all over the world instead of one server hosting the movie or the or the website data, a thousand servers might host it. Eventually, there will be a layer on top of that called file coin, which actually incentivizes people to store all that data all over the world. So it's a little bit like mining by hosting that data, they get a little bit of reward of file coin and what you end up with is essentially like a decentralized data center that stores all this data using people's hard drives all over the world instead of using centralized servers. So the non-incentivized version is called IPFS, the incentivized version is called file coin. And both of those are really interesting technologies that we use them in some of our architectures.
CCN: And I guess the other interesting thing you could do there, you and I both download the Ethereum white paper to our laptops. We uploaded to Google Drive and they're storing two copies, right? So if you guys built a retail-based application, you and I would be storing or referencing the same object.
Sam Cassat: Yeah, yeah, exactly. Content addressable is a very interesting property of it.
CCN: And how is the database lookups down there? Right? Because we're talking about potentially hundreds of thousands of files.
Sam Cassat: It's been a while since I've delved into the...
CCN: Oh yeah, yeah. No problem.
Sam Cassat: ...guts of IPFS is engineering, but essentially it's a search across the network. I mean, you know, all of these nodes, store all this data. There's some registry that's decentralized where everything is and a message gets sent out like a beacon looking for the data. Eventually, the data gets sent back after that beacon was sent out essentially.
CCN: So that's actually a really interesting point. ConsenSys has 50 projects?
Sam Cassat: Fifty, a little under 50 portfolio companies in our venture studio. In total -- we have a lot of projects. But yeah, about 50 things in various states of development in our venture studio.
CCN: Then you have the consulting arm, a dozen other things. How as an executive at this company, do you keep up with that and how as the chief strategy officer of such a decentralized type company do you steer the ship?
Sam Cassat: ConsenSys is a really unusual organization, right? We're decentralized. What that means, is different in different contexts. But in a sense when I make a decision or if I think something is a good idea, it doesn't necessarily mean I can tell someone to do it. We're more about the wisdom of the crowds. We're more about crowdsourcing. We're more about reaching consensus, which happens to be part of our moniker. So really we have a lot of really brilliant people that work for us.
When we want something to happen, it usually results in discussion. If I think something is a good idea or I think something needs to happen then usually what -- the way that we make that happen is you get a whole bunch of interested stakeholders and parties together, all of which are probably pretty smart people and have a discussion about it and reach a consensus on what to do and then that kind of filters throughout the organization. It's a very large organization. We have a lot of companywide meetings, we have a lot of communication. We have Slack with an extremely large amount of information. We have a couple of meetings that are for the whole company every two weeks and every three weeks. So if you want to, you can let them know everything that's going on. We have an event probably almost every day around the world. So just having really high touch, really large amounts of communication. And really open dialogue is, I think, the best way to surface the wisdom of the really smart crowd that is within ConsenSys. So it's a little different than a normal company, but I think it also, it allows us to surface more information. And it allows us to move really in the sort of spirit of what we're doing, which is building revolutionary technology that changes the world. And if we didn't try to do something different we would be charlatans, I think. We're trying to walk the walk and make our whole company and experiment. It's been a pretty successful one so far, I think.
CCN: So just out of curiosity, I'm trying to understand how the company's governed and how it works. Let's say your consulting arm has a giant client that wants to use file coin. And at this point file coin is decentralized and you need to make a pretty big architectural change there. That means you need to have the consensus of the community, the miners and all these different stakeholders. How do you go about communicating that, working with the developers and all the different stakeholders to achieve consensus and move forward at a pace that an enterprise type client would be used to from like an Oracle?
Sam Cassat: They're two different things going on. There's the public Ethereum network. Which would have file coin running on it and be a completely decentralized architecture. And then there are private implementations Ethereum and private Blockchains or semiprivate consortium Blockchains for maybe several companies share those projects and that infrastructure. If it's something where it's a private context and maybe there are ten banks sharing it or there are ten oil and gas companies sharing it or something, it's pretty easy for us to modify anything you want and just push that project forward and do whatever we want. I don't think the market is quite there yet where these big corporates and governments, at least for the most part, are putting their projects on public Ethereum infrastructure.
I think we will get there. I think we'll get to a point just like we are right now with the normal technology market where the market will demand certain functionality of the Ethereum Blockchain ecosystem.
Players will then utilize that. Those participants in that technology available in the market to build what they want. Those things will meet each other but right now I wouldn't say -- we wouldn't try to go cram something into the market, but it wasn't ready for. But in the private implementations, we can do whatever we want.
CCN: So as far as distributed ledger technology, what do you think is the most compelling use case at the moment? Whether you guys are working on something or not?
Sam Cassat: Uh, the most compelling use case? Well, I think the obvious use case, the obvious killer app that has already been the killer app is tokenization. We've already disrupted the venture capital industry. We've already, we've already changed the face of how startups grow and fund and capitalize themselves. So that's a pretty killer app. I think security tokenization, like Meridio, this shirt I'm wearing I think is a really great application. I think we have yet to see even the very beginning of how much innovation will come in the financial sphere from the fact that a software object with programmable properties can also be a financial product and a security and the legal framework around that will support that behavior. I think we're one step down a long road of that and then there's tremendous innovation there.
I think providing access to asset classes and to finance in general for emerging markets is a tremendously powerful use case. I think disrupting journalism as civil as doing is a ripe area for decentralization. Right now we're, we're seeing the battle play out between the centralized forces of our government and Facebook and Google and Twitter. And we had Jack and Cheryl in Congress yesterday testifying about whether and how their companies should censor and change the news or fake news or whatever. Our society is at a moment where decentralization might make more sense than centralization on those fronts. And we're grappling with that right now. I think that's a really ripe area. There are a lot of them.
CCN: And ConsenSys is certainly working on tackling them all. So I'd really like to thank you for coming on and it was awesome talking to you.
Sam Cassat: Thank you. Really cool to be here. Thank you.
Note: This interview is part of the CCN Podcast. The podcast and this interview are also available on iTunes, TuneIn, Stitcher, Google Play Music, Spotify, SoundCloud, YouTube or wherever you get your podcasts. Make sure you rate and subscribe!
Last modified (UTC): September 16, 2018 3:35 PM