Joseph Lubin could arguably be seen as a star at any blockchain event these days, but at the Ethereal Summit last month — organized by Ethereum-focused development company ConsenSys, which he co-founded — he was especially revered.
While inside the conference space, the co-founder of Ethereum was either listening intently to what was happening on stage or surrounded by a group of people between panels. Nevertheless, he found some time to speak with us during the two-day event.
So, on a sunny Saturday, we sat down on a quaint cast-iron bench in the garden area of the Brooklyn venue, Pioneer Works.
Lubin spoke about the current technical capacity of Ethereum and what awaits the ecosystem in the near future. Our conversation touched on scalability, consensus protocols and public vs. private blockchains for businesses looking to integrate the technology.
This interview has been edited and condensed.
Can Ethereum scale?
Olivia Capozzalo: Yesterday, Jing from Plasma Group spoke very enthusiastically about Ethereum's ability to scale. Can you comment on that? Can Ethereum scale? And if so, what are the main developments right now in the ecosystem that are making that happen?
Joseph Lubin: So, I think the point, as she punctuated that panel quite brilliantly and entertainingly, was not so much about its ability to scale, it's that it has already scaled quite significantly.
So, she's part of the Plasma Group. It's a group that is pioneering a class of different solutions for scalability. Essentially, recognizing that we have this base trust layer that can handle 15 to 27 transactions per second. And above that we have state channels of various different varieties in zk-STARKs and in zk-SNARKs and Truebit and Plasma.
And Plasma is this class of technologies that enable you to have less decentralized platforms sitting at layer two in the Ethereum ecosystem. They can benefit from the full trust in some cases — sometimes they benefit from partial trust — but if they're linked in really rigorously, they can benefit from the full trust of the base trust layer, and you can get the best of both worlds.
Watch the full interview with Joe Lubin:
So, you get high transaction throughput per second, and you get the security of the base layer. And by that I mean, if you have a game and you brought your own network for your game or your exchange or some other application, if they have assets on your system, everybody using your system can be confident that, if you're incompetent or if you're corrupt, it doesn't matter so much. It's a pain in the butt, but they can still pull your value tokens back to safety and you're not vulnerable. So, that's happening.
So, I think we're at many tens of thousands of decentralized transactions per second on the Ethereum network right now. And another point that I believe she was making, and that I think Ameen was making, is that we've got all this scalability for specific use cases.
So, we're not reaching any limits soon with the base trust layer at 15 to 27 transactions per second, but the base trust layer within 18 to 24 months is going to multiply its capacity by about a thousand times.
That development is Serenity, or Ethereum 2.0. And Ethereum 2.0 is divided into four phases: three major phases — and in computer science terms, they're numbered 0, 1 and 2. Phase 0 is getting close, it's eight different groups that are building their own clients according to a specification that's really very stable right now, a bunch of different testnets that each uses and there's one testnet for everybody to come together.
So, within a small number of months, we should have a fully operational testnet and possibly by the end of this year, we'll have a fully operational real Phase 0 Ethereum 2.0. So, good chance it'll go live this year.
There'll be different ways that it gets connected with Ethereum 1.0: ether tokens will move from Ethereum 1.0 to Ethereum 2.0, there may be bidirectional mechanisms, and there may be a way in the not too distant future to use the Beacon Chain, which is basically the Phase 0 proof-of-stake finalized blocks on Ethereum 1.0.
A proof-of-stake future
OC: Okay, so you mentioned proof-of-stake and I wanted to ask about another point from yesterday that was sort of contentious with the panel about proof-of-stake vs. proof-of-work. I know Ryan Selkis from Messari was sort of critiquing proof-of-stake.
JL: So, I don't know that he was critiquing proof-of-stake. The question that was put to him about having a certain amount of money to invest in either Bitcoin or Ethereum 2.0. He said that he would put 80% or 100% on Bitcoin, because Ethereum 2.0 isn't released yet. There're still questions and he has children.
I don't think he was fully discounting proof-of-stake, I think he just knows that it has been proven that proof-of-work works.
And so, if he was faced with the conservative decision of investing his child's college fund, he would make the prudent choice. That is kind of the choice we made on the Ethereum project at the start, we intended to go proof-of-stake.
Unlike what was said on that panel, there are proof-of-stake systems that are working — different flavors of proof-of-stake systems. But we were aware of edge cases in all of the systems that were working at the time that could potentially take down a network.
Those systems weren't incredibly valuable, and we figured that if the Ethereum network is incredibly valuable, then well-resourced actors would potentially exploit these vulnerabilities.
So, we knew we could make proof-of-work work, and the intention was to do that, do the research and get to the point where we were very confident in proof-of-stake — and that is done.
OC: Could you summarize why it’s so important to move to proof-of-stake?
JL: Proof-of-work is a mechanism that keeps all the different nodes of a network in sync. So, it's a consensus formation mechanism. You get the trust characteristic from blockchain, from having all these nodes kept in sync.
Proof-of-work is one class of consensus algorithms — they all, essentially, find a leader and everybody follows in behind that leader. And this is a brand new one, it's a decentralized mechanism where you don't really elect a leader, the leader wins its role and wins the right to propose the next state of the system by showing everybody the next block that's valid. Then, everybody validates that and crypto economics causes them to all fall into sync.
But proof-of-work, unfortunately, requires very expensive custom hardware, enormous amounts of electricity and wastes huge amounts of computation, and it benefits efficiencies of scale. So, if you're a well-resourced actor, you can have an unfair advantage over someone running it on their game machine at home.
Proof-of-stake fixes all of that, proof-of-stake trades all that expense for a crypto-economic bond — essentially ether [ETH] — that you put into a smart contract on Ethereum. It burns orders of magnitude, less electricity, so you'll be able to run it on your pad or phone at some point pretty soon, or some jewelry at some point in the not-too-distant future.
It doesn't waste a lot of computation. It has very low barriers to entry, so my sister could do it or somebody could set up a warehouse, and my sister wouldn't be disadvantaged compared to that warehouse — because, essentially, it's probabilistic in terms of how much you're called on to participate, depending on how much you've invested.
So, it's a more secure system and a fairer system — more equitable. Because it's based on probabilistically selecting validators for each new block, you can have a single validator pool for many different sharded blockchains. Right now, we have a single validator pool that keeps Ethereum's blockchain secure, so all the validation power is focused on that one blockchain.
Split all that validation power into 1,024 different shards that would weaken all the different shards and people would notice that shard number 37 is really weak and these other shards would gang up and it would be madness. But from this one validator pool in Ethereum 2.0, groups are selected and randomly allocated to validate the different shards, so all the shards are secured equally and they're all secured with the full validation power of the entire network.
Ethereum for enterprise
OC: I also wanted to touch on the debate between public and private blockchains. As we heard on another panel, representatives from EY and ConsenSys were both arguing for using Ethereum's mainnet, a public blockchain, for large enterprises.
JL: We do a huge amount of work in our solutions group and we've built lots of enterprise blockchains, private permissioned blockchains for companies and for consortia, and banks and central banks. And you absolutely need to build the right architecture for each use case. There aren't a lot of use cases on public blockchains right now that are appropriate for enterprise use cases, enterprise solutions.
One of [ConsenSys’] John Wolpert's arguments is that Ethereum will be the base trust layer, the base settlement layer that many different sidechains and other technologies will link into. We've got a group called Aztec that built a protocol that enables obfuscation of transactions on the public Ethereum.
The Aztec protocol is super cool, and it will be live on public Ethereum pretty soon. That's very similar to what Ernst & Young [EY] built, so [EY’s] Paul Brody described Nightfall, which also enables the shielding of public transactions on public blockchain.
Essentially, the public Ethereum isn't fully ready for primetime — for all use cases — because it's not scalable enough yet and because it doesn't have sufficient privacy and confidentiality for all use cases yet.
We're solving privacy and confidentiality by using private networks that can link into the public Ethereum or link into each other. We're also solving it with those two protocols that I just described, so many different classes of transactions or use cases can now, or soon, be done on the public Ethereum. And we're solving scalability via layer two and moving to Ethereum 2.0.
OC: Awesome, that's really great. Thank you so much, really appreciate it.