On March 23, I hosted a Twitter Spaces discussion with U.K.-based Dan Hughes, founder of Web3 startup Radix DLT and a respected cryptographer who spent years single-handedly designing his own version of blockchain technology. He talked through the technical challenges of Web3, and why it could take up to a decade to go mainstream.
Parmy Olson: So let’s just start by going back to 2008, when the famous Bitcoin white paper by Satoshi Nakamoto came out. That paper essentially sparked what has become a global industry centered around the concept of Bitcoin and other cryptocurrencies. You read that paper and immediately started working on a crypto-related project that could do a better job of scaling. What was so important about it and what was it that it showed everybody for the first time?
Dan Hughes: For me, it showed a solution to a long-debated problem in computer science, which was how do you arrive at consensus when you have a permissionless population of validators in the network. As a tech nerd, that was the first thing that really made me go, “Wow.” The solution that Satoshi arrived at is really quite elegant and covers all the bases. It wasn’t until a little bit later that I really started to think about the possible implications of this technology and what it could mean for us as a species.
PO: When people today talk about Web3, how much are they talking about those same principles that were raised in that white paper more than 13 years ago?
DH: The same principles still apply. Web3 will still very much be grounded in those first principles that Satoshi laid out in the Bitcoin white paper. Without them, Web3 just isn’t possible and you’d just end up with a kind of Frankenstein Web 2.0 that doesn’t really maximize the potential that you could otherwise.
PO: What is the big hope here for Web 3? What could it do for the world that our current internet, or Web 2.0, doesn’t do?
DH: For me, the crux of Web 3 for me is that it offers choice. Up to quite recently, you didn’t have a great deal of choice as an individual. The obvious example is in the financial system. Prior to Bitcoin, if you weren’t desirable to the financial system and you couldn’t get a bank account then you were kind of stuck. You were very limited in the choices that you could make around how to manage your money, how you could invest that money, how you could try and maximize the value of that money by starting a business and stuff like that.
Now we have a lot of choice in terms of our money. We have Bitcoin and an endless number of other cryptocurrencies, and I can choose to win yield on it, I can lend out to people, I can start a business within the DeFi ecosystem that’s currently evolving. Web 3 is an extension of that choice. If you are content creator, say you’re making videos, then you don’t have a lot of choice of where to go with your content. It’s mainly YouTube or Twitch and once it’s there, you’re at the mercy of their policies. Web 3 allows you to have a lot of choice over a broader spectrum of your life.
PO: You’re building the infrastructure, and one of the things you’re doing with Radix is designing this single protocol for scaling up decentralized finance. What does that mean?
DH: Let’s say that tomorrow, everybody on planet Earth wakes up and wants to get involved in DeFi. The first issue you’ve got is, “How do you support that level of volume?” All of those transactions and actions being made by seven billion people — that’s potentially millions of actions per second. If you don’t have a core protocol that all of your DeFi applications can live on top of that can handle that volume of throughput and user interactions, then you’re going to trip as soon as you get off the starting block.
PO: Is it the case that there isn’t there isn’t a single protocol, but lots of different protocols and is that a problem?
DH: I don’t think it’s a critical problem. I think that if you have too many protocols, then there’s a lot of complexity that can arise there because they’ve all got to be able to communicate with and send tokens to each other. There’s a lot of friction that adds a lot of latency, which can be detrimental. So I’m not kind of advocating that there should only be one protocol, but if you start with the mentality of, “Well, what if there was only one protocol, what would it require?” then you’re future-proofing from the start, instead of trying to retroactively improve it later on.
PO: Speaking about different protocols, if we wanted Web 3 to work as seamlessly as Web 2.0, would there need to just be a few protocols or one standardized protocol for different things that everybody is using? How does that compare to how Web 2.0 works?
DH: Over the long term, I think you’ll definitely see a bunch of clear winners that carry the bulk of Web 3 traffic and interactions. In the very early stages of the internet 25 years ago, there was a lot of different protocols such as TCP/IP, Gopher, FTP and they all served a different purpose. There wasn’t really one unified protocol that everything was built on top of, it evolved into that instead. I think you will find a similar thing happens here.
PO: Let me put the big question to you. What do you think are the biggest challenges that could stop Web 3 from going mainstream?
DH: We’ve touched on one, which is protocols working together. But you really need to look at it from a user perspective. If it’s too difficult or risky to use, then people just aren’t going to engage with it. The majority of people involved in this space at the moment are quite tech-savvy and willing to take on risk. But when you are talking to Bill the plumber in the local pub, he isn’t really interested in understanding all of the technical complexity that is currently required to be able to use this stuff. He doesn’t want to have to worry about the risk of it being stolen or losing his keys. There’s things like regulation as well. If we don’t work with the regulators to make sure that these regulations don’t overreach, then that can also be a nail in the coffin here too.
PO: To be fair, technology regulators have tended not to overreach with things like social media. But I want to go back to your point about user experience, which resonates for me as someone who’s tried using distributed apps. I tried using a Twitter alternative called Peepeth a couple of years ago. Instead of posting a tweet you posted a a “peep,” and each time you wanted to do that, you had to use a certain number of tokens using your crypto wallet, and it took a day for a peep to actually show up on the network. I found it clunky to use. Do you think this usability issue is why we’re not seeing other apps like this?
DH: There’s a recurring issue here underneath a lot of these things. It fundamentally comes down to the scalability properties that the particular protocol can handle. If you were to move everybody off Twitter onto some alternative Web 3 platform then it would just grind to a halt. The technology just isn’t there to be able to support these super large-scale, big data applications that we’re used to using as Web 2.0 consumers every day. But that will get solved over time. The other issue is, of course, it is clunky and difficult to use. You have to send tokens constantly for the things that you want to be able to post or interact with on these platforms. Part of the reason for that is actually because the development experience is also very immature and quite difficult.
Coding a smart contract in (Web 3 programming language) Solidity is not easy. There are a lot of “gotchas” that you have to be aware of so that you can avoid them. We see all the time that certain products have been subject to a hack or a bug in the code that’s meant $1 million has been stolen by somebody, for example. Most of these exploits that happen on smart contracts are because of one or two lines of code where whoever was developing in Solidity was maybe having a bad day and just overlooked something, or didn’t understand all of the security concerns that you need to take into account. Something as simple as a typo can cause these issues which can be exploited by savvy hackers. So as a developer, there’s a real learning curve at the moment to be able to build applications that are secure and function well.
PO: So actually programming distributed apps is harder compared to Web 2.0 because if you make a mistake with a Web 3 app, you’re much more likely to have a gaping security hole or a bug?
DH: It’s a lot more brittle. You also have to consider the fact that these apps are dealing with generally large sums of money. If I was building a decentralized Twitter application, and there was a bug which meant some tweets went missing or some pictures didn’t upload properly, it’s a mild annoyance. But if I make a mistake in a (cryptocurrency exchange) Uniswap-type application, and real money goes missing, like people’s life savings, then there’s a lot of pressure there, right? It wouldn’t be too bad that the development environment is brittle, if there wasn’t all this responsibility for people’s funds as well.
PO: I want to bring up this blog post that was written in January by Moxie Marlinspike, the famed cryptographer who created Signal and the encryption underpinning Signal and WhatsApp. His big critique was that Web 3 services just aren’t as decentralized as everybody makes out. How big a problem is it if Web 3 is actually pretty centralized?
DH: At the moment, Web 3 is a very young concept. A lot of the bits and pieces still need to be built. So you could have a truly decentralized (NFT marketplace) OpenSea, it just hasn’t been built yet. For at least the short, medium term, a lot of Web 3 services will be very centralized. They will use the blockchain as a means to keep an ultimate record of what’s happened, but will be accessed via these kind of centralized gateways. But as the space matures, you will definitely find that a lot of these services start to become more decentralized.
If we rewind five or six years, there were no decentralized exchanges at all. But you look today and every platform has got multiple decentralized exchanges. The reason for that is that it takes time to build this technology and test them and make sure that they’re truly working in a decentralized, secure way.
On the other hand, centralized services aren’t always bad, they do have their place. The key thing missing here is that people aren’t informed and educated enough on what are the pros and cons of both of them are. The one thing that I do think is important is that the underlying protocols themselves must be decentralized at all times.
PO: Why do the underlying protocols themselves have to be decentralized?
DH: Because the decentralization is in part a catalyst for the security of that system. So if it was very centralized, then you run a higher risk and a potential of single actors being able to control that protocol and do things like censor transactions or prevent you from trading or creating content or selling your NFT. So centralization of a core protocol just leads to a large number of potential security risks and issues that you might run into.
PO: It does sound like some of Web 3’s biggest challenges are under the hood. How big a deal are those issues and how much do you think investors and entrepreneurs need to be aware of them?
DH: Right now? Not so much. As an industry, we’re still finding our way. There are still a lot of hard technological issues to solve and there’s a lot of potential in a lot of places. But just be aware, this stuff is gonna take time.
PH: Can I ask you a really annoying question, Dan? Can you guesstimate when Web 3 might go mainstream?
DH: My gut feel is five to 10 years. And by “mainstream,” I’m thinking a billion users. This space is moving extremely fast. I’ve been part of a lot of cutting edge technologies in my time and this is by far the fastest moving space I’ve ever been involved in. A year here is 10 years everywhere else.
PO: Speaking as someone who has been in this game for around a decade now, when you first started coding and trying to build your own version of blockchain, that was moving very slowly. Has this industry only been moving quickly in the last year?
DH : So building protocols that could support the kind of throughput and scale that we need for this is a very hard problem and is just going to take however long it takes. But we’ve come a long way already in the past 10 years, and I’ve no doubt in the next couple of years, we’ll actually see protocols that will be able to handle this kind of load. Then the other side of the coin, pardon my pun, is you need all of the tools and the ability to build front ends that can interact seamlessly with them.
You’ve got a multitude of things that you need to all fall into place before you can really even start to think about pushing this out to the mainstream. But it’s moving fast. Uniswap wasn’t even a thing a few years ago, and now we’ve got Uniswap and complex lending tools and farms and yields and pool swaps. It’s moving very fast now because the fundamental components are being put in place ever more quickly.
More From Bloomberg Opinion:
• They May Be Kids, But the Lapsus$ Hackers Are Still Giants: Tim Culpan
• Will Russians Use Crypto to Bust Sanctions?: Editorial
• Web3 Just Had Its Emperors-New-Clothes Moment: Parmy Olson
This column does not necessarily reflect the opinion of the editorial board or Bloomberg LP and its owners.
Parmy Olson is a Bloomberg Opinion columnist covering technology. She previously reported for the Wall Street Journal and Forbes and is the author of “We Are Anonymous.”
