This month Blockstack’s “Decentralizing the World” tour made its final stop in Hong Kong. At the event, top technologists gathered to discuss topics such as new models of governance, digital rights, venture capital in crypto, the Stacks Blockchain, and more.
Nick Grossman of Union Square Ventures presented on some of the inherent tensions and unique challenges to finding product market fit while building on blockchain systems. He then breaks them down into three primary groups; Control, User Experience, and Layers. How this generation of decentralized projects deals with these challenges will determine how successful they are and factor mightily into how fast they can achieve more mainstream adoption.
Control
Companies have the luxury of total control of their product and ecosystem. This is not the case for decentralized protocols. Once a project becomes decentralized enough, it becomes nearly impossible to accomplish the high speed pivots that are sometimes necessary in the course of managing a company. So, how does a decentralized project strike a balance between working toward decentralization while still shepherding it in the right direction and responding to changing circumstances?
User Experience
The lack of control and clean governance (when compared to traditional companies) means that decentralized protocols have an uphill battle when trying to foster clean user experiences, which are typically borne of a tight vision. That can make initial adoption more difficult, or at least something that has to be managed more closely. An upside is that the apps built on these decentralized protocols will have more user customizable experiences given their decentralized (and often open source) nature. For example, if you use Evernote and don’t like a recent change, you’re stuck with it. In a decentralized ecosystem, you would own your data so you can simply take your notes to a new client that does things how you like or build your own.
Layers
Protocols can be layered on top of each other, with each level building on the security and features (or lack thereof) of the chain below it. For example, Blockstack is built using the most secure open blockchain at the moment, Bitcoin. Because these protocols are open source and permissionless, anyone can experiment with combining them in interesting ways. Hopefully, by encouraging and funding this kind of experimentation, we will eventually land on on more and more right answers to the first two points.
For more, watch Nick’s full presentation above or read through the transcript below. If you’re looking to attend a Blockstack event in the near future, please visit our community calendar. You can also view the other talks and presentations from Hong Kong here.
Full transcript of Nick’s talk:
00:04 It’s wonderful to be here in Hong Kong and at the Hong Kong Exchange in a global financial center thinking about a new kind of money, crypto-currencies, that are also the underpinning for a new kind of computing: decentralized computing.
00:21 And the way that those two things come together is really fascinating. I will say that today’s conference has been an incredible stage setting for the big ideas here. Why is this important? Why do we care? And I wanna talk a little bit more practically about how do we get there, ’cause I think it’s important to work on that.
00:37 And so really, what I wanna talk about is finding product market fit in decentralized apps and Web 3. We’ve got the big ideas, we think we know why we’re doing this, but how do we build things that consumers and businesses really want? How do we iterate? It’s not exactly obvious how we do that, it’s a little bit harder in this space than it might be in some other spaces.
01:03 I’ve been thinking a little bit about this question in how do you find product market fit, which is do you start with decentralization and work your way towards market fit, or do you start by finding market fit and then work your way towards something more decentralized? And we’re seeing projects take both approaches. I’m not here to say which is the right approach, it’s just a question that’s been on my mind.
01:26 And I wanna put a picture in your minds, this is a graph of global smartphone adoption. The orange is the iPhone and blue is Android. And what you notice is at the beginning, it was only the iPhones, only iOS and it was dominant. Why was it dominant? Because Apple, because they controlled the entire thing top to bottom, Apple’s a dictatorship. It’s not a democracy, it’s closed, it’s not open.
01:51 They were able to find fit, control, everything they wanted to do, iterate, experiment and build a very compelling product and win hearts and minds with that product. But what happened over time? Over time, the Android approach to having a more open eco-system, the open source approach is by far the more scalable approach.
02:09 And now Android has six or seven X the market share of IOS. And so, I think it’s a great example of this interplay between closed and open, centralized, decentralized, what makes it easier to innovate, what makes it harder to innovate at the beginning and over time?
02:26 So that’s a little bit of a backdrop. Just quickly who I am, I’m from Union Square Ventures, we’re a venture capital firm based in New York City, we’ve been around since about 2004. We’re known for making a lot of investments in internet applications, social networks like Twitter, marketplaces like Etsy, developer tools like MongoDB and Twilio.
02:45 Today, much of our focus is in education and healthcare. Back in 2012, we started investing in the crypto space because we see this as the next generation of the platform that everything’s gonna get built on.
02:59 For reasons that we’ve been talking about all day today, mostly about trust and integrity. So we invested in coin-based back in 2013, and have backed a small handful of projects building new protocols like Blockstack, Algorand, Filecoin and a handful of others.
03:16 And so first, why are we all here today? Why are we having this conversation? Why do we care about decentralization? It’s an interesting question, and it’s a little bit like the elephant and the blind man with their hand on the elephant. Everybody when we talk about decentralization or crypto, has a bag of ideas that they think it is, but it’s really hard to get the whole picture.
03:40 And depending on where you’re coming from, it looks like something very different. But we know that it’s about certain big ideas like trust and integrity and control, self-sovereignty, permissionless innovation which was at the heart of the internet and we’ve lost a fair bit of.
04:00 But above all, it’s about something new. I think we are at the beginning of creating something brand new, and having a platform that allows us to create lots of new things that we can’t even imagine yet.
04:13 And so I don’t even think we’re looking at an elephant, maybe it’s a platypus with a duck bill and otter feet and a wide tail, but I think it’s really interesting because we just don’t exactly know, we feel like we have the piece but we’re not exactly sure where it’s gonna go yet.
04:29 And the process of iteration and experimentation and innovation and finding product market fit for all of these platforms and protocols and apps is important to me, because it’s how we’re gonna get from the big ideas that we’re talking about today, to millions of people using this technology in the next five or 10 or 15 years.
04:51 Just on the platypus real quick, one example that I think about sometimes is the creation of Linux. So when Lenox was made and to a large degree it was a response to Windows, and the Linux desktop, a lot of effort went into it. But the Linux desktop was not the product that was transformative for Linux, it was the Linux server that enabled the internet.
05:15 And the same way that I think it’s very important to re-architect the web as we know it today, and I think that’s gonna happen, but I also think we’re gonna lay the ground work for something brand new that we can’t even imagine yet.
05:28 A couple of challenges that are unique to this space and this moment in this space. The first one is that companies are not the same as protocols. Companies like startups or companies like Amazon and Google. In a company, it’s a dictatorship.
05:42 You can decide what to do, you can move fast and break things, it’s about agility and change, and change at a moment’s notice. And that is part of why developers in particular are wary of building on top of corporate platforms, because the rules can get changed at any minute. It’s why we like building on the internet, ’cause it’s more stable.
06:04 And then you have protocols. Protocols are enduring, they’re stable, they’re slow. They change generally speaking, not through a centralized autocratic process, but through a decentralized governance process. Governance in air quotes because that can mean a lot of different things. And even the fastest governance process is still very, very slow compared to the corporate process.
06:26 And in the distributed apps in crypto space, we’re sort of at this early stage where it’s a little bit like needs to be agile company development, but we’re also building these long-term enduring protocols that are hard to change. And so that creates for a complex moment and a challenge in finding product market fit.
06:45 Just to illustrate this, this is the stack of protocols on the internet, the hourglass. And IP being the protocol of the middle, at the physical layer, everything compiles up to IP. And then above that, it spreads out, and there’s lots of different protocols. At each layer as you get away from IP, there’s more variety.
07:05 And the color-coding is to say, the protocols that matter the most, that are at the center like internet protocol are the hardest to change. They’re widely distributed, they change very slowly. In fact, they really shouldn’t change very often and very quickly.
07:20 If you look at how long it’s taking to move from IPv4 to IPv6, it’s a decade-long process. Whereas, applications on top of this stack, are in an incredible position. Because of this layered architecture and because of the openness of it, and because of the ability to build on top of whatever protocols you want, you can iterate on applications very fast.
07:40 And that’s why there’s been so much amazing innovation in startup land on top of the internet. But that has started to change as the applications at the top have gotten thicker, platforms like Google, Apple, Facebook Amazon. And for application developers, they’re no longer building on top of an open protocol stack, they’re building on top of a protocol stack that they don’t control. And so that’s what’s so interesting about where we are right now, and I think why developers are among the first folks who are, have a strong feeling about why all this matters.
08:10 The second interesting challenge is the tension between decentralization and user experience. It’s not to say that they are fundamentally opposed, but they are often in tension. So if you think about why is there Coinbase on top of Bitcoin, because sometimes you need to have a smooth, controlled user experience that’s harder to do in a decentralized environment.
08:35 Or why do we have Spotify when there was BitTorrent? It’s because Spotify is easier to use and there’s a smoother experience there. But it’s not always that centralized sense of a better user experience, it’s just that it’s easier to iterate towards a better user experience.
08:54 But creating good user experiences for people is fundamental to getting this technology adopted. So how do we do that? There’s a few approaches, and this gets back to decentralizing first or later. The first is do less with more, and by more I mean more security and more decentralization.
09:11 You can have a narrower product, you can have a blockchain that’s slower that has higher security, a higher decentralization, or you could do more with less. A lot of projects are trying to make the foundational platform faster with a trade off against security and decentralization.
09:31 And it’ll be very interesting to see which of those approaches matters in the short-term or works in the short-term and also in the long-term. Another way of saying that is whether projects starting out in a space should be experience first or decentralization first.
09:50 And so getting to where does that leave you if you’re working on a platform or if you building decentralized application? How should you think about all this, and what are some possible routes to take? I think the first one is sort of obvious, but it’s going where decentralization is the feature.
10:10 And that’s not necessarily true for every application, but I think it’s particularly true for some applications in, for instance, frontier markets where there is no incumbent, is no legacy and there is no infrastructure, or places where censorship resistance is really a killer feature and is a unique feature.
10:29 It’s an interesting question to think about how many of those use cases are there and how big are they? And can we build products and applications that solve them in a way that’s so much better that accounts for the challenges in building a good user experience.
10:44 Another approach is to decentralize some of the things, to be surgical about where you apply decentralization and how you use it. I think the blockstack approach is a really interesting one. They’re not trying to put everything on chain, the goal is not to decentralize, or to put everything on chain. The goal is to surgically use the highest security component for identity, and then decentralize in a different way off-chain. And that is I think a very smart approach that gets you the highest value benefit from certain types of decentralization in certain places. I think we’ll see more projects start to do that as we go on.
11:31 Another approach, which is I think controversial but also worth talking about, is decentralizing gradually. Taking a little sip, seeing how it goes and going farther. A couple of examples, another one of our portfolio companies is Numerai, which is an online data science tournament backed by a token. It’s an ERC20 Ethereum token.
11:52 And really it began as a website that used the token, and the token was part of the mechanics of running the website. But it wasn’t fully decentralized in ways that most people would recognize, but having the token in the website enabled the team to really iterate on the process. And then in the fall, they started launching this thing called Erasure which is taking what all of the centralized aspects of Numerai and the Numerai token, and decentralizing them.
12:25 Taking away control from the folks who had been at the center, taking the hands off the wheel and handing it over and having it become more of a fully autonomous smart contract as opposed to a web application that used a token.
12:37 I think that’s very interesting, and I’m not sure that the Numerai and Numerai IR team would’ve gotten to where they’re getting with Erasure if they had gone fully decentralized from day one. So we’ll see, it hasn’t launched yet but that’s where they’re heading.
12:53 Another one is CryptoKitties, another one of our portfolio companies. Anybody who plays CryptoKitties might realize that the smart contract itself is on the Ethereum blockchain, but the artwork lives on CryptoKitties website. So it’s decentralized ownership record of a very centralized asset. If their company goes out of business, the artwork disappears.
13:16 And so what they’re doing now is, they’re working on putting the artwork itself into IPFS, which is a decentralized file system, decentralized addressing and storage system. So moving from a centralized model to a more decentralized model.
13:29 I will note that the reason that they’re able to do this, is that Cloudflare, another USD portfolio company that operates a very centralized CDM and security network with global footprint, is running IPFS nodes so that this peer to peer network of content addressing and IPFS is globally available.
13:51 So back to the idea of the pendulum between closed and open and centralized and decentralized, I think we will continue to see certain decentralized elements interfacing with certain centralized elements in order to create the right balance of control and integrity and trust. That’s the thing that we’re really working towards.
14:12 The last thing is probably the most important thing, is about layers. What’s the most exciting thing to me about crypto and decentralized applications is the ability to build freely and permissionlessly on top of open layers. Back to the diagram about the internet protocols earlier, that is the thing that made the internet interesting. It’s the thing that enabled experimentation on a broader scale than we’ve ever seen before. And I think we’re coming back to that with distributed applications, decentralized applications and crypto networks.
14:46 A friend of ours, Dennis whose startup we backed and is now Andreessen Horowitz, describes it as following which is that every blockchain and smart contract is defacto API. Everything is open by default. Not only is it forkable and copyable, but it’s buildable. And that’s what’s happening with blockstack too.
15:05 The thing that I’m most excited about today with blockstack and the idea of the personal data storage is that I can have layers and layers of applications building on top of my data, once that’s open and under my control, I’m in an open system.
15:19 And also, blockstack is another good example of layering on top of Bitcoin, and this is a graph of the lightning network traffic since launch basically; and it’s growing. And this is consistent with, whether or not this is the right thing or not long-term, the idea of building on open layers that are all protocols, that are all non-proprietary and that are powered by some sort of cryptocurrency is probably the most powerful thing that’s going to unlock experimentation in this category that’s gonna get us to the point where we have more product market fit for more products.
15:58 And I think over time, we’ll end up with something that looks like a protocol stack, like we have with the internet. It was sort of a rational, established protocol stack, not a chaotic, messy protocol stack like we have today. But we’re very, very early today in trying to understand what are the best protocols for what purposes? How do they fit together? But we are clearly in the business of building the next layer of protocols that will enable a very special kind of innovation that we haven’t seen in quite some time.
16:31 The only thing I know for sure, is that it won’t be clean and neat. It will be an elephant fight for dominance of core protocols and we’re already seeing that. It’s gonna be confusing, it’s going to be fun. So that’s all I have today, thank you so much for having me, great to be here and I look forward to on this with all of you.