S1:E4 Ethereum Foundation's Justin Drake–The Merge, POS and Where ETH Will Be In 50 Years

Katherine Wu

Katherine [00:00:04]

Hello everyone and welcome to episode four of Cross-Chain Examination. I'm your host, Katherine Wu.

So for anyone who's either new to crypto or is an expert, Ethereum is probably one of the most important ecosystems. That's where a lot of companies are built and really where smart contracts originated. So obviously, given its importance in the crypto ecosystem, the fact that Ethereum is about to undergo a major transition is probably one of the most important topics to understand today.

Today's guest is Justin Drake, who honestly, I feel really grateful to have convinced to be on the show. Justin is probably one of the very few leading developers working on Ethereum. He's done this for the past five years at the Ethereum Foundation, and Ash and I use this opportunity today to ask him all things Ethereum front and center.

Okay. Hello everyone and welcome to episode four of Cross-Chain Examination. Our big topic today is all things Ethereum, specifically looking forward to the summer and digging into the upcoming merge, which is the big transition for Ethereum that many have been waiting for. As Ethereum enters its last chapters of a proof of work system, we're going to talk about what The Merge means, what to expect, and how it works. For today's episode, we actually also have on a special co-host Ash Egan who was actually our guest just two episodes ago.

Ash [00:02:07]

Hi, everyone. Ash here, we're really excited about today's episode. We're welcoming our guest, Justin Drake, who's been a researcher at the Ethereum Foundation for the last five years I believe - Justin correct me if I'm wrong - and welcome to the show, Justin. So good to have you.

Justin [00:02:23]

Yeah, thanks for having me.

Katherine [00:02:25]

So to give a quick lay of the land: Ethereum, as we know it, is about to transition from a proof of work system to a proof of stake system. And that's probably the highest level TLDR on The Merge. And Justin is our expert here. So I'm going to turn it over to Justin to tell us what exactly is The Merge and how should the everyday users think about The Merge.

Justin [00:02:49]

Right. So as you said, The Merge is basically a migration from proof of work to proof of stake. And it's kind of a white elephant in the sense that it has a number of advantages and it depends on your vantage points, whether you appreciate the advantages. If you're kind of a mainstream person, maybe one of the striking benefits is that Ethereum will no longer be using the amount of electricity of a small country. So that's a big win. I guess if you're slightly more specialized and less mainstream, you might be a gamer. And in that case, you will be happy to hear that the price of graphics cards should go down. So basically all the millions of graphics cards that are currently being used to secure Ethereum will return to the market and that should provide some relief there for gamers.

Katherine [00:03:46]

That's the good news. We're all happy to hear that. Can you actually tell us why it's even called The Merge?

Justin [00:03:53]

The reason why it's called The Merge is that in order to deploy proof of stake safely, we actually had kind of a dress rehearsal period for proof of stake. And the way that we did that is that we have two chains running in parallel. We have the main chain, which some people used to call Ethereum 1, which is based on proof of work. And it's what secures the EVM and all the transactions that run on Ethereum and all the smart contracts. And in parallel, there's the so-called Beacon Chain, which is the proof of stake chain, which has been running for about a year and a half. So it's been well tested and it currently has about $25 billion of Eth staked. So a high amount of lindy from a point of view of value security. And so we're at a point where we're reasonably confident or not very confident that the proof of stake chain can secure the EVM and all those transactions. And so these two chains rallying in parallel will merge into a single chain. And in the process we will be shedding away the proof of work layer and replacing it with proof of stake.

Ash [00:05:05]

Incredible. Lots of exciting stuff going into The Merge obviously. I'm sure it's a pretty stressful environment over at the EF and you know, sort of helping with this transition. I think one narrative that we've been seeing more of in the last few weeks and months is, and especially as it relates to proof of stake, is sort of this growing market share of stakers. You look at LIDO and their percentage of the network that they're validating. Can you just sort of give us a high level on what your thoughts are on that centralization, just given how high of a priority decentralization has been for Ethereum since its early days?

Justin [00:05:47]

Basically the product that Ethereum is selling is secure block space. And so really anything that's related to security is of interest to us. Now it turns out that an attacker that's large enough on Ethereum can do certain bad things. It's very limited in what it can do, but it can do some bad things. But let me just start with things that an attacker can’t do. And by an attacker, a large attacker, I mean a majority attacker with more than 50% of the Eth staked.

One of the things that they can’t do is just change the rules of Ethereum. So it can't just mint Eth out of thin air, for example. And another thing that it can't do is roll back blocks past a finality point. So we have this finality gadget, which every epoch which is 6.4 minutes, is going to finalize a new epoch. And an attacker can't basically revert the chain past that.

But a large attacker does have some power and it's basically related to censorship and short reorgs past the last finality points. And this is bad for multiple reasons. Like one of the reasons is that the attacker might be just purely malicious and they want to disable the Ethereum chain. And one way to do that is just to produce all empty blocks and just completely destroy the liveness of Ethereum. Another thing that such a large attacker could do is basically try and extract as much money as possible. And this is where MEV comes in and maybe perform reorgs that are several blocks deep and replay transactions in an order that favors them. And there may be like a final reason why a large attacker might attack Ethereum is just to be compliant with regulation. As you could imagine, nation states that ask certain addresses to be frozen and blocked. And this is kind of considered an attack on this firm because it goes against the credibility and neutrality of this area.

Now, the good news is that for all these flavors of attacks, we have a fallback, which is that the social layer, what we call social consensus. So the social layer can come together, observe the attack, and thanks to proof of stake, the consensus participants are acting in broad daylight. And so we can identify them as a community and address the situation. But there’s kind of two subsequent questions that we want to ask ourselves. The first one is how can we minimize the probability that we get into this bad situation to start with? And this is where decentralization is important, because if you have one big entity that controls a large portion of staking, then that could be, for example, a regulated entity. And that could be a pressure point for governance. And then the other reason why it's important to have decentralization is that if and when social governance comes in and neuters the attacker, there needs to be something to fall back to. And here, basically, we want to lower the barrier to entry so that at minimum, we kind of have these very robust, very decentralized home stakers that are distributed across many many different jurisdictions. And that will kind of act as a fallback. And so this is why it's important that we shouldn't have too much centralization at a minimum. It would be great if we had, let's say, 10 million Eth that is staked at home. And if we have 10 million Eth that is staked at home and that is immune to all these shenanigans, then basically Ethereum can only be attacked on the order of ten times. And the reason is that every time the social layer comes in it’s going to slash or destroy or burn at least 10 million Eth and because it's only on the order of 100 million Eth, that attack can only happen ten times.

Ash [00:09:52]

Maybe we can just double click on the staking providers in particular as it relates to centralization and decentralization. I hope we live in a future where you have folks at home that are running validators, 32 Eth via Raspberry Pi or whatever it is. And that’s actually something that we're looking in the market for and looking to make that more accessible for the average ETH holder. On the contrary, we have seen somewhat of a cartelization of these staking providers. I think your framework is terrific, but maybe just to double click on that, what are your views on the future of everyone staking from home or democratizing staking in general and decentralizing it in effect?

Justin [00:10:41]

So I don't think everyone will be staking at home. But I think it's sufficient that a large enough minority of people stake from home. But I think what is important is that there isn't too much centralization. And especially we don't want to go past what we call protocol thresholds. So there are certain attacks that become unlocked past a certain threshold. So the one third or 33% threshold is an important one. The 50% threshold or one half is very important as well. And then there's the ⅔ thresholds. And then depending on where you are on the threshold, you can perform different types of attacks. And really where we want to be in the position ideally is that no single entity is above this one threshold. I mean, an even better situation would be like even the top two largest entities combined don't make up 33%. Now it's very, very early days of Ethereum staking and so there isn't that much choice and diversity for users. But the underlying staking is extremely malleable. You can write all sorts of wrapper contracts and you can create all sorts of so-called liquid staking derivatives (LSDs). And so what I expect will happen is that there will be lots of choice and going forward, more decentralization. Right now, LIDO has about, you know, 30% of all the stake. One of the things that people don't realize is that lives are always controlled by a governance token - the LDO token. And the market cap of the LDO token is on the order of a $300-$350 million. If an attacker can get 50% of the LDO token, they can compromise all the staked if under little. But there's about $10 billion of staked ETH. And so the attacker has a huge amount of leverage. You know, they only need to invest a few hundred million dollars in order to compromise $10 billion of staking.

So I think what will happen is several things. One is that the existing services like LIDO will harden and in time will become more decentralized themselves because right now they are quite centralized with points of failure. And so these will be patched and I think there will also be more competition. And also another thing that's going to happen is that the barrier to entry to becoming a solo validator will go down. Just to give you one barrier to entry. After The Merge, the fees and the MEV is going to accrue to the validators. Now it turns out that if you're just one single validator running from home, you're only going to make a block once every few months. And there's going to be a huge amount of variance. And so really what the validators want is so-called MEV smoothing. They want to be in a position where they get a consistent amount of revenue as opposed to basically buying lottery tickets and gambling. And this is the service that pools can provide. But this is something that we will provide in-protocol with basically enshrined MEV smoothing. Another big barrier to entry for solo stakers is the minimum amount of ETH that you need to stake, right? It's 32 ETH, but the technology is being built right now so that multiple small home validators can pull together in a trustless way. For example, 32 different individuals each putting down one ETH and then combined making up one one validator.

Katherine [00:14:28]

Going back to the earlier question about The Merge and also what you've just talked about. So the goal of moving to a proof of stake network - aside from lowering the barriers to entry to run your own validators, you kind of enable decentralization that way, you're maybe more environmentally friendly - are the bigger or some of the bigger narratives around why proof of stake? It should enable cheaper gas fees or maybe faster transactions because there's less congestion. In your opinion, in the post Merge world, where do the layer 2s or the Ethereum's scaling solutions then sit? So for example like Arbitrum or Optimism is there a future where they can coexist or do they become competitive with Ethereum?

Justin [00:15:16]

So I guess to finish answering your first question around the advantages of The Merge, there's kind of two advantages that we haven't touched upon. One is basically what I call economic efficiency, and that boils down to how much issuance does the system need to have to pay for one unit of security and specifically one unit of economic security. So the way that we measure the security of a blockchain is through economic security, which is basically the the number of dollars that an attacker needs to have in order to overwhelm the consensus and perform a 51% attack. And the economic efficiency for proof of stake is on the order of 10 or 20x better than proof of work. And so what that means is that we can drastically reduce the amount of issuance. And kind of as a secondary benefit to that we're actually going to be in a position where the ETH supply is going to go down over time. And the reason is that we have a very healthy amount of burn every single day and deal with a 5000ETH or 4000ETH, which is burned every single day. And right now with proof of work, we have to issue 13,500 ETH every single day, which is larger than what we burn, but with proof of stake, we're bringing that way down to 2000 ETH. And so we'll have a decreasing supply. And I think this is a narrative that eventually will be appealing to the kind of mainstream investors people who watch CNBC. But at this point in time, it hasn't really percolated all the way to the mainstream.

And then the final advantage, and for me it's the most interesting one, but it's kind of the geekiest and the hardest to appreciate, is that it's an improvement in security. And it's an improvement for two reasons. One is that it allows us to have much more economic security. That's largely due to the fact that the economic efficiency is so much better than proof of work. So right now, if you want to attack Ethereum, you need to spend on the order of $10 million to overwhelm consensus with hashrate. So you need to buy $10 million worth of hardware and cooling and transformers and rackspace and all of that stuff. With proof of stake today, even in its very nascent early days, it's $25 billion, and I expect that to grow to 100s of billions of dollars, if not over a trillion or trillions of dollars eventually. So it's going to be way more secure than alternative proof of work systems like Bitcoin even. The other big advantage which was touched on is that it allows for social consensus to intervene in case that there is an attack. Again, this is a feature which proof of work doesn't have. If an attacker can attack proof of work, they can do so on a repeated basis. The term that we use is a spawn camp attack and it comes from gaming. In gaming, when you shoot someone down, they respawn at a specific location. And one of the strategies that you can take is that you put all your men right where you respawn and then you just keep attacking the respawned players. Now, what was your question again?

Katherine [00:18:44]

It was about layer 2s, but as a right. But I actually really think it's so funny how so much of gaming influences crypto. If you read Vitalik’s origin story, he was attracted to the idea of unstoppable or things that can be whatever he was just attracted to the idea of decentralization because of Warcraft. He lost his game where he lost his in-game assets or money or something. It's just funny.

Justin [00:19:15]

His warlock was nerfed, which I think is technical language for it, it was less powerful or something like that. Right.

Katherine [00:19:23]

Forgive my crude description.

Justin [00:19:27]

Okay. So scaling. So proof of stake doesn't immediately provide scaling relief. And that's kind of one of the common misconceptions is that at The Merge, fees will go down. They won't go down. If anything, they'll probably go up. The reason being that fees tend to be correlated with the price of ETH. And if there's a lot of excitement around The Merge, price goes up, then fees will go up. But you mentioned the rollout. So yeah, we kind of have this two pronged strategy to scaling. One is basically providing more resources at the consensus layer. And then the second part of the strategy is being able to consume these resources that the blockchain provides in a more clever way.

So the first part is called sharding. And the idea here is to provide on the order of 100x more scaling. And the way that we do that is that instead of asking every single consensus participant to validate the whole blockchain, we instead ask consensus participants to validate small parts of the chain in a randomly selected fashion. So we kind of use randomness to select smaller committees, for example, that perform specialized tasks.

And then there's this other aspect of scaling, which is making better use of the resources that we have. And this is called rollups, and that provides another 100x. And so these two 100x combined provide a 10,000x. So if we have ten transactions per second on Ethereum today for the whole world, which is definitely not enough, we will be in a position to have 100,000 transactions per second in a couple of years because that's the timelines for both rollups and sharding combined.

Now the way that rollups work is basically the realization that to have consensus over execution, it suffices to have consensus over data. So that's a bit of a technical mouthful, but the idea is that when you consume Ethereum as a resource, you're doing two things. One is that you're putting the transaction data on-chain and here you're consuming bandwidth effectively. And the other thing that you can do is you can execute these transactions, you can interpret them and modify the state. And here you're using resources like CPUs and disk space and RAM. And the very clever realization is that it turns out you only need to spend data, and data is quite a bit cheaper than execution on the order of 100 times. And so we can just stuff all this data on-chain which represents many more transactions and we don't execute them. Instead, the trick is - let's execute them off chain. And there's basically two flavors of rollups. There’s the optimistic rollups, whereby you do the execution off chain, people make claims about what the outcome of these executions are, if these claims are wrong, anyone can trustlessly challenge them, and they guarantee to win the challenge. So it's kind of a crypto economic game. And then the second favor is called ZK rollups, where using clever cryptography called snarks, basically math proofs, that the execution was done correctly so that you don't have to do the full brute force naive execution of transactions. Instead, the blockchain can just verify this very succinct proof on chain. The S in snarks stands for succint.

Ash [00:23:05]

Maybe just to dive into that a little bit more and just play out a scenario. So as it relates to these scaling, all the scaling approaches obviously mainnet rollups, let's say I'm a college student, I'm studying computer science, I'm fascinated by what's happening in Web3, I want to build decentralized Uber. If you were giving me advice, would you say go deploy that on a rollup? Go wait a few years? Deploy it on main? Deploy it across a few different rollups and mainnet. How would you think about that for the folks who are thinking around building applications for the mainstream?

Justin [00:23:49]

Right. So I would definitely encourage you to deploy on a rollup. And the reason is that very soon there's going to be a mass migration of users from L1 to L2. The reason being that general purpose rollups like Optimism and Arbitrum, which are live, and we now have incentives for people to move over to this new chain. So for example, the Optimism change just released the OP token and hopefully these two things combined will be sufficient to migrate network effects. And once you've migrated, there's really no reason to go back to L1 because you're just going to be paying 10 to 100 times higher transaction fees. There is this question around which flavor to deploy to - optimistic or ZK roll up. Well, today you have no choice. You have to use an optimistic rollup. But I think ultimately ZK rollups will win. But from the point of view of a developer, that's kind of a detail, right? Because a lot of these rollups are trying to target the EVM. So just build as if you're building on the EVM and then the infrastructure and the back end will improve over time.

So there's kind of two ways in which you can benefit from ZK rollups. Way number one would be for an optimistic rollup to upgrade to a ZK rollup. And I think this is the endgame for basically all optimistic rollups. Or maybe more plausible is that one of the ZK rollups manages to gain network effects. And the reason why I say more plausible is that you need to have a very, very strong technical team with very specialized knowledge in order to deploy one of these ZK rollups, so these dedicated teams will have a first mover advantage. So if you want to take a bit of a risk, then you can maybe try and bet on one of these ZK rollup teams, wait a few months until they're ready, and deploy there.

Katherine [00:26:00]

So now that we've covered a lot about The Merge, Ethereum, what it means move to proof of stake. I kind of wanted to switch gears and actually ask about the community behind Ethereum. So one of which is the Ethereum foundation, which is Justin where you're doing research. Can you tell us more about the Ethereum Foundation as an organization? So how it's involved with developing Ethereum or the Ethereum ecosystem? And just kind of tell us more about the foundation, about your role and how it really pushes technological innovation on Ethereum.

Justin [00:26:37]

So the Ethereum foundation is a not for profit, which is funded partly from the Bitcoin sale in 2014, I believe, and also from the adaptation of ETH that it received at the genesis of Ethereum. And basically its remit is to push forward the R&D of Ethereum, especially the R more than the D, the research. And also push forward public goods and fund them through grants. And so the foundation today has grown quite large. It has about 260 people. It's split up in many different teams that do different things. One of the teams, for example, is the grant giving team. And so here, basically what they do is process grant applications and try to find the best way to allocate a budget on the order of tens of millions of dollars every single year towards public goods that will support the ecosystem.

There's also various teams like the Solidity team and Geth team and the Socrates team, which is basically building core infrastructure for Ethereum consensus or peripheral infrastructure to support Ethereum. And then the team that I'm part of, which is the research team, which is, of size, roughly 20 people. What we do is try and use crypto economics, so basically both cryptography and incentives, to design systems that fulfill the criteria that we want around security and around scalability. And what we do is, the end product, is basically written specs. So we have these executable specs written in Python. And then the interesting thing is that these specs don't get implemented internally. It's a different foundation. Instead, what happens is that we have teams all around the world that do the heavy lifting of implementing these specs. It’s somewhat similar to the IETF, which is kind of a spec writing body for the internet. And then various teams like Google, you know, for Google Chrome, Firefox, Edge, all implementing these internet protocols in the context of a browser. And so one of the things that kind of comes out of this setup is client diversity. So we have, for example, five different teams that have been writing five different implementations in five different languages of the proof of stake spec.

And in addition to the research team, we also have a security team and a testing team. So one of the things that we found is that there are some benefits to centralization within the EF (Ethereum Foundation), specifically when it comes to security quality assurance, because it's difficult for every single implemented team to have security experts and to have testers. And so one of the things that we've been building in anticipation of the match is this security team internally to the EF that looks at all of the implementations in one go.

And so really, if you kind of zoom out, there's a bit of a pipeline, right? So, many years ago Vitalik planting the seeds of a vision, right? Even before Ethereum was launched, even before Genesis, Vitalik envisioned proof of stake and sharding. And then the goal of the research team is kind of to dream up these crypto economic games and then write them down and formalize them as a spec. And then that moves onto the implementer teams which are spread all across the world and have different sources of funding. And then it kind of comes back a little bit towards the EF for QA, audits, and things like that. And now when it comes to The Merge we’re at the very, very later stages basically of polishing and QA. All the heavy lifting has been done and in particular the research has been completely finalized. So one of the things you said at the beginning of the podcast is that it must be very stressful at the EF. Well for me I've done my work when it comes to The Merge, at least from the research standpoint.

Ash [00:31:19]

Now you're just memeing.

Justin [00:31:22]

Just memeing. Also helping coordination around QA and things like that.

Ash [00:31:28]

It's remarkable Vitalik seeing around corners, having a vision of the future everything you've done, the quality of folks at the EF. Just really incredible to see what you guys have done in such a huge lift in transitioning. There's so many doubters. It's never going to happen. This shift to proof of stake. So t's fun to see it's right around the corner.

Maybe just as a final question, and we talk about this a little bit in person a few weeks ago at Permissionless, but what is Ethereum in 10 years? In 50 years? How should we as participants think around what Ethereum will become or could become?

Justin [00:32:15]

Let's start with what Ethereum could be in ten years. So one of my theses is around shared security. This is the idea that there are network effects for applications being built upon a common platform, a common secure platform. If we have a blockchain which is simultaneously extremely high grade from a security standpoint and can scale to the whole world, then the network effects of such a platform would mean that it's kind of a winner take most. And so there is a possibility that in ten years, let's say, Ethereum becomes the substrate for securing the internet of value, basically.

Now, you may ask, hold on, the scalability roadmap that you put forward gives us 100,000 transactions per second. Is that really sufficient for the whole world? Now it turns out that if you allow yourself to project ten years into the future, there is another benefit. Scalability benefits basically another 100x. And the way that you achieve another 100x is by looking at the equivalent of Moore's Law for bandwidth. So that's a law called Nielsen’s Law. And it basically states that the consumer bandwidth grows 50% every single year. And this is something that we should expect to continue for at least a decade because bandwidth is this massively paralyzable thing. And even on this single strand of fiber, you can pass insane amounts of data - petabytes and petabytes and petabytes per second.

And it turns out that if you look at the fundamentals of scaling consensus, you need to look at the computational resources at hand, right? There's bandwidth, there's CPU, there's disc IO, which is basically reading and writing to disk, which is currently the bottleneck and there's storage. And three of those four computational resources can be removed as a bottleneck. Let me give you the example of CPU. How do you remove CPU as a bottleneck? The way that you do that is you use zero knowledge proofs. And so one of the things that we want to do on Ethereum is to basically snarkify the EVM, to create a ZK EVM in such a way that in order to become a validator, you only need to check these tiny proofs. You don't have to spend all the computation. And so that removes CPU as one of the bottlenecks for becoming a validator.

And then there's this other upgrade called statelessness, which removes two of our bottlenecks, namely the storage and the disc IO. And what are you left with? You're left with the fundamental computational constraint for scaling, which is bandwidth. And that, thanks to Nelisen’s law, is going to grow 100x over the next ten years. And so we're going to be in a position where we have 10 million transactions a second, which is more than enough for everyone on Earth.

I guess in ten years time, we can also make a prediction around the supply of ETH. So right now there's 120 million ETH and the supply might go down by let's say roughly 2 million ETH every single year. And so in ten years time, we should be around 100 million ETH in total circulation. And the fact that Ethereum kind of stands out as being the only chain right now that is basically capable of pulling such a feat of basically being able to extract more value in the form of the burn than value that it consumes just to secure itself and basically have this deflationary aspect to its supply may mean that ether the asset could become money for the internet. So basically we have this yin and yang where we have Ethereum, the platform, the blockchain, which secures the Internet of value. And then we have ether, the asset, which is optimized to be a collateral asset used in the context of staking by placing ETH as collateral to become a staker, but also used in the context of Defi, for example, in the case of decentralized stablecoins.

Now I guess the other part of the question was, what could be Ethereum in 50 years? And I guess there's kind of a couple of outcomes here. One is that it's become boring, right? It's become so pervasive and so commoditized within society that it's just this thing which is totally well understood. It's not changing. It’s totally immutable. And it's securing the internet of value. And that's kind of the happy case. The not so happy cases, that’s in 50 years is just completely forgotten and the relevance and there’s maybe a better piece of technology or some sort of competitor that takes over.

If we take the happy case that we can also look at what the supply of ether will be. What will happen over the long term is that the supply will find an equilibrium very similar to Bitcoin's 21 million limit. There will be an equilibrium for ether. And this is basically dictated by what percentage of the circulating supply is burned every single year. So if there's 1% of the circulating supply, which is burned every year, then the equilibrium is going to be 100 million ETH. If it’s 2% it’s going to be 50 million ETH. If it’s 3% is going to be 33.3 million ETH. My expectation is that in the endgame is that there's going to be  roughly 60 million ETH in circulation. And so just like Bitcoin kind of finds this equilibrium over decades towards 21 million through inflation, right? Bitcoin will be inflating for the next century, basically. It will be the very similar thing for ether the asset, but in the other direction. It will be deflating for 50 to 100 years and then eventually find this equilibrium, which could be, let's say, around 60 million. And so basically there's going to be a halvening, not of the issuance well of the supply, today we have 120 million ETH, and that could have to 60 million ETH over the very long term.

Ash [00:39:16]

And worst case, we end at 69 million ETH.

Justin [00:39:21]

Well, we do know that whatever happens, there will be some sort of equilibrium that that's the expected thing. So one of the things that Bitcoiners like to say is that the supply of ether is infinite, but if there is some kind of percentage of the circulating supply, which is burned every single year, then that gives us an equilibrium, whatever happens. And so that could be 200 million ETH, it could be higher than what it is today or it could be lower than what it is today.

Katherine [00:39:48]

Well, Justin, thank you so much for coming on today. We talked about a lot. We talked about The Merge. We talked about proof of stake. We talked about the future of Ethereum. And I think these are all really important topics to follow. Last question before you go - where can listeners go and follow the work that your team at the EF does or in general keep up with developments happening in the Ethereum ecosystem?

Justin [00:40:10]

Most of the researchers on Twitter, I'm @drakefjustin there. We also post on ETH research and in general, what tends to happen is that a lot of us work from home and then once in a while we go to a big conference. And one of the reasons why we go to a conference is not for the conference per say, it’s because immediately before or immediately after the conference we have some sort of research retreat. And so if you want to meet us a lot of us will be going, for example, to EthCC in Paris and a lot of us will be coming to DEFCON in Bogota. So, yeah, I'd love to meet you in person.

Katherine [00:40:51]

Awesome. Well, thank you so much, Justin. And we'll have you again. Some point again.

Justin [00:40:57]

Thanks for having me. Thanks, guys.

Katherine [00:41:01]

Thank you everyone for tuning in to episode four of Cross-Chain Examination. Please like and subscribe wherever you're getting this podcast, whether it's YouTube, Apple Podcasts, Spotify or the like. Leave me any thoughts, questions, suggestions for either this episode, past episodes or future episodes. Thank you again and I'll see you next week.

accelerating the decentralized future

Ep.1: Crypto Coven–Building An Inclusive And Sustainable NFT Community

July 6, 2022
 | 
45:09
 | 
S1:E4

Katherine [00:00:04]

Hello everyone and welcome to episode four of Cross-Chain Examination. I'm your host, Katherine Wu.

So for anyone who's either new to crypto or is an expert, Ethereum is probably one of the most important ecosystems. That's where a lot of companies are built and really where smart contracts originated. So obviously, given its importance in the crypto ecosystem, the fact that Ethereum is about to undergo a major transition is probably one of the most important topics to understand today.

Today's guest is Justin Drake, who honestly, I feel really grateful to have convinced to be on the show. Justin is probably one of the very few leading developers working on Ethereum. He's done this for the past five years at the Ethereum Foundation, and Ash and I use this opportunity today to ask him all things Ethereum front and center.

Okay. Hello everyone and welcome to episode four of Cross-Chain Examination. Our big topic today is all things Ethereum, specifically looking forward to the summer and digging into the upcoming merge, which is the big transition for Ethereum that many have been waiting for. As Ethereum enters its last chapters of a proof of work system, we're going to talk about what The Merge means, what to expect, and how it works. For today's episode, we actually also have on a special co-host Ash Egan who was actually our guest just two episodes ago.

Ash [00:02:07]

Hi, everyone. Ash here, we're really excited about today's episode. We're welcoming our guest, Justin Drake, who's been a researcher at the Ethereum Foundation for the last five years I believe - Justin correct me if I'm wrong - and welcome to the show, Justin. So good to have you.

Justin [00:02:23]

Yeah, thanks for having me.

Katherine [00:02:25]

So to give a quick lay of the land: Ethereum, as we know it, is about to transition from a proof of work system to a proof of stake system. And that's probably the highest level TLDR on The Merge. And Justin is our expert here. So I'm going to turn it over to Justin to tell us what exactly is The Merge and how should the everyday users think about The Merge.

Justin [00:02:49]

Right. So as you said, The Merge is basically a migration from proof of work to proof of stake. And it's kind of a white elephant in the sense that it has a number of advantages and it depends on your vantage points, whether you appreciate the advantages. If you're kind of a mainstream person, maybe one of the striking benefits is that Ethereum will no longer be using the amount of electricity of a small country. So that's a big win. I guess if you're slightly more specialized and less mainstream, you might be a gamer. And in that case, you will be happy to hear that the price of graphics cards should go down. So basically all the millions of graphics cards that are currently being used to secure Ethereum will return to the market and that should provide some relief there for gamers.

Katherine [00:03:46]

That's the good news. We're all happy to hear that. Can you actually tell us why it's even called The Merge?

Justin [00:03:53]

The reason why it's called The Merge is that in order to deploy proof of stake safely, we actually had kind of a dress rehearsal period for proof of stake. And the way that we did that is that we have two chains running in parallel. We have the main chain, which some people used to call Ethereum 1, which is based on proof of work. And it's what secures the EVM and all the transactions that run on Ethereum and all the smart contracts. And in parallel, there's the so-called Beacon Chain, which is the proof of stake chain, which has been running for about a year and a half. So it's been well tested and it currently has about $25 billion of Eth staked. So a high amount of lindy from a point of view of value security. And so we're at a point where we're reasonably confident or not very confident that the proof of stake chain can secure the EVM and all those transactions. And so these two chains rallying in parallel will merge into a single chain. And in the process we will be shedding away the proof of work layer and replacing it with proof of stake.

Ash [00:05:05]

Incredible. Lots of exciting stuff going into The Merge obviously. I'm sure it's a pretty stressful environment over at the EF and you know, sort of helping with this transition. I think one narrative that we've been seeing more of in the last few weeks and months is, and especially as it relates to proof of stake, is sort of this growing market share of stakers. You look at LIDO and their percentage of the network that they're validating. Can you just sort of give us a high level on what your thoughts are on that centralization, just given how high of a priority decentralization has been for Ethereum since its early days?

Justin [00:05:47]

Basically the product that Ethereum is selling is secure block space. And so really anything that's related to security is of interest to us. Now it turns out that an attacker that's large enough on Ethereum can do certain bad things. It's very limited in what it can do, but it can do some bad things. But let me just start with things that an attacker can’t do. And by an attacker, a large attacker, I mean a majority attacker with more than 50% of the Eth staked.

One of the things that they can’t do is just change the rules of Ethereum. So it can't just mint Eth out of thin air, for example. And another thing that it can't do is roll back blocks past a finality point. So we have this finality gadget, which every epoch which is 6.4 minutes, is going to finalize a new epoch. And an attacker can't basically revert the chain past that.

But a large attacker does have some power and it's basically related to censorship and short reorgs past the last finality points. And this is bad for multiple reasons. Like one of the reasons is that the attacker might be just purely malicious and they want to disable the Ethereum chain. And one way to do that is just to produce all empty blocks and just completely destroy the liveness of Ethereum. Another thing that such a large attacker could do is basically try and extract as much money as possible. And this is where MEV comes in and maybe perform reorgs that are several blocks deep and replay transactions in an order that favors them. And there may be like a final reason why a large attacker might attack Ethereum is just to be compliant with regulation. As you could imagine, nation states that ask certain addresses to be frozen and blocked. And this is kind of considered an attack on this firm because it goes against the credibility and neutrality of this area.

Now, the good news is that for all these flavors of attacks, we have a fallback, which is that the social layer, what we call social consensus. So the social layer can come together, observe the attack, and thanks to proof of stake, the consensus participants are acting in broad daylight. And so we can identify them as a community and address the situation. But there’s kind of two subsequent questions that we want to ask ourselves. The first one is how can we minimize the probability that we get into this bad situation to start with? And this is where decentralization is important, because if you have one big entity that controls a large portion of staking, then that could be, for example, a regulated entity. And that could be a pressure point for governance. And then the other reason why it's important to have decentralization is that if and when social governance comes in and neuters the attacker, there needs to be something to fall back to. And here, basically, we want to lower the barrier to entry so that at minimum, we kind of have these very robust, very decentralized home stakers that are distributed across many many different jurisdictions. And that will kind of act as a fallback. And so this is why it's important that we shouldn't have too much centralization at a minimum. It would be great if we had, let's say, 10 million Eth that is staked at home. And if we have 10 million Eth that is staked at home and that is immune to all these shenanigans, then basically Ethereum can only be attacked on the order of ten times. And the reason is that every time the social layer comes in it’s going to slash or destroy or burn at least 10 million Eth and because it's only on the order of 100 million Eth, that attack can only happen ten times.

Ash [00:09:52]

Maybe we can just double click on the staking providers in particular as it relates to centralization and decentralization. I hope we live in a future where you have folks at home that are running validators, 32 Eth via Raspberry Pi or whatever it is. And that’s actually something that we're looking in the market for and looking to make that more accessible for the average ETH holder. On the contrary, we have seen somewhat of a cartelization of these staking providers. I think your framework is terrific, but maybe just to double click on that, what are your views on the future of everyone staking from home or democratizing staking in general and decentralizing it in effect?

Justin [00:10:41]

So I don't think everyone will be staking at home. But I think it's sufficient that a large enough minority of people stake from home. But I think what is important is that there isn't too much centralization. And especially we don't want to go past what we call protocol thresholds. So there are certain attacks that become unlocked past a certain threshold. So the one third or 33% threshold is an important one. The 50% threshold or one half is very important as well. And then there's the ⅔ thresholds. And then depending on where you are on the threshold, you can perform different types of attacks. And really where we want to be in the position ideally is that no single entity is above this one threshold. I mean, an even better situation would be like even the top two largest entities combined don't make up 33%. Now it's very, very early days of Ethereum staking and so there isn't that much choice and diversity for users. But the underlying staking is extremely malleable. You can write all sorts of wrapper contracts and you can create all sorts of so-called liquid staking derivatives (LSDs). And so what I expect will happen is that there will be lots of choice and going forward, more decentralization. Right now, LIDO has about, you know, 30% of all the stake. One of the things that people don't realize is that lives are always controlled by a governance token - the LDO token. And the market cap of the LDO token is on the order of a $300-$350 million. If an attacker can get 50% of the LDO token, they can compromise all the staked if under little. But there's about $10 billion of staked ETH. And so the attacker has a huge amount of leverage. You know, they only need to invest a few hundred million dollars in order to compromise $10 billion of staking.

So I think what will happen is several things. One is that the existing services like LIDO will harden and in time will become more decentralized themselves because right now they are quite centralized with points of failure. And so these will be patched and I think there will also be more competition. And also another thing that's going to happen is that the barrier to entry to becoming a solo validator will go down. Just to give you one barrier to entry. After The Merge, the fees and the MEV is going to accrue to the validators. Now it turns out that if you're just one single validator running from home, you're only going to make a block once every few months. And there's going to be a huge amount of variance. And so really what the validators want is so-called MEV smoothing. They want to be in a position where they get a consistent amount of revenue as opposed to basically buying lottery tickets and gambling. And this is the service that pools can provide. But this is something that we will provide in-protocol with basically enshrined MEV smoothing. Another big barrier to entry for solo stakers is the minimum amount of ETH that you need to stake, right? It's 32 ETH, but the technology is being built right now so that multiple small home validators can pull together in a trustless way. For example, 32 different individuals each putting down one ETH and then combined making up one one validator.

Katherine [00:14:28]

Going back to the earlier question about The Merge and also what you've just talked about. So the goal of moving to a proof of stake network - aside from lowering the barriers to entry to run your own validators, you kind of enable decentralization that way, you're maybe more environmentally friendly - are the bigger or some of the bigger narratives around why proof of stake? It should enable cheaper gas fees or maybe faster transactions because there's less congestion. In your opinion, in the post Merge world, where do the layer 2s or the Ethereum's scaling solutions then sit? So for example like Arbitrum or Optimism is there a future where they can coexist or do they become competitive with Ethereum?

Justin [00:15:16]

So I guess to finish answering your first question around the advantages of The Merge, there's kind of two advantages that we haven't touched upon. One is basically what I call economic efficiency, and that boils down to how much issuance does the system need to have to pay for one unit of security and specifically one unit of economic security. So the way that we measure the security of a blockchain is through economic security, which is basically the the number of dollars that an attacker needs to have in order to overwhelm the consensus and perform a 51% attack. And the economic efficiency for proof of stake is on the order of 10 or 20x better than proof of work. And so what that means is that we can drastically reduce the amount of issuance. And kind of as a secondary benefit to that we're actually going to be in a position where the ETH supply is going to go down over time. And the reason is that we have a very healthy amount of burn every single day and deal with a 5000ETH or 4000ETH, which is burned every single day. And right now with proof of work, we have to issue 13,500 ETH every single day, which is larger than what we burn, but with proof of stake, we're bringing that way down to 2000 ETH. And so we'll have a decreasing supply. And I think this is a narrative that eventually will be appealing to the kind of mainstream investors people who watch CNBC. But at this point in time, it hasn't really percolated all the way to the mainstream.

And then the final advantage, and for me it's the most interesting one, but it's kind of the geekiest and the hardest to appreciate, is that it's an improvement in security. And it's an improvement for two reasons. One is that it allows us to have much more economic security. That's largely due to the fact that the economic efficiency is so much better than proof of work. So right now, if you want to attack Ethereum, you need to spend on the order of $10 million to overwhelm consensus with hashrate. So you need to buy $10 million worth of hardware and cooling and transformers and rackspace and all of that stuff. With proof of stake today, even in its very nascent early days, it's $25 billion, and I expect that to grow to 100s of billions of dollars, if not over a trillion or trillions of dollars eventually. So it's going to be way more secure than alternative proof of work systems like Bitcoin even. The other big advantage which was touched on is that it allows for social consensus to intervene in case that there is an attack. Again, this is a feature which proof of work doesn't have. If an attacker can attack proof of work, they can do so on a repeated basis. The term that we use is a spawn camp attack and it comes from gaming. In gaming, when you shoot someone down, they respawn at a specific location. And one of the strategies that you can take is that you put all your men right where you respawn and then you just keep attacking the respawned players. Now, what was your question again?

Katherine [00:18:44]

It was about layer 2s, but as a right. But I actually really think it's so funny how so much of gaming influences crypto. If you read Vitalik’s origin story, he was attracted to the idea of unstoppable or things that can be whatever he was just attracted to the idea of decentralization because of Warcraft. He lost his game where he lost his in-game assets or money or something. It's just funny.

Justin [00:19:15]

His warlock was nerfed, which I think is technical language for it, it was less powerful or something like that. Right.

Katherine [00:19:23]

Forgive my crude description.

Justin [00:19:27]

Okay. So scaling. So proof of stake doesn't immediately provide scaling relief. And that's kind of one of the common misconceptions is that at The Merge, fees will go down. They won't go down. If anything, they'll probably go up. The reason being that fees tend to be correlated with the price of ETH. And if there's a lot of excitement around The Merge, price goes up, then fees will go up. But you mentioned the rollout. So yeah, we kind of have this two pronged strategy to scaling. One is basically providing more resources at the consensus layer. And then the second part of the strategy is being able to consume these resources that the blockchain provides in a more clever way.

So the first part is called sharding. And the idea here is to provide on the order of 100x more scaling. And the way that we do that is that instead of asking every single consensus participant to validate the whole blockchain, we instead ask consensus participants to validate small parts of the chain in a randomly selected fashion. So we kind of use randomness to select smaller committees, for example, that perform specialized tasks.

And then there's this other aspect of scaling, which is making better use of the resources that we have. And this is called rollups, and that provides another 100x. And so these two 100x combined provide a 10,000x. So if we have ten transactions per second on Ethereum today for the whole world, which is definitely not enough, we will be in a position to have 100,000 transactions per second in a couple of years because that's the timelines for both rollups and sharding combined.

Now the way that rollups work is basically the realization that to have consensus over execution, it suffices to have consensus over data. So that's a bit of a technical mouthful, but the idea is that when you consume Ethereum as a resource, you're doing two things. One is that you're putting the transaction data on-chain and here you're consuming bandwidth effectively. And the other thing that you can do is you can execute these transactions, you can interpret them and modify the state. And here you're using resources like CPUs and disk space and RAM. And the very clever realization is that it turns out you only need to spend data, and data is quite a bit cheaper than execution on the order of 100 times. And so we can just stuff all this data on-chain which represents many more transactions and we don't execute them. Instead, the trick is - let's execute them off chain. And there's basically two flavors of rollups. There’s the optimistic rollups, whereby you do the execution off chain, people make claims about what the outcome of these executions are, if these claims are wrong, anyone can trustlessly challenge them, and they guarantee to win the challenge. So it's kind of a crypto economic game. And then the second favor is called ZK rollups, where using clever cryptography called snarks, basically math proofs, that the execution was done correctly so that you don't have to do the full brute force naive execution of transactions. Instead, the blockchain can just verify this very succinct proof on chain. The S in snarks stands for succint.

Ash [00:23:05]

Maybe just to dive into that a little bit more and just play out a scenario. So as it relates to these scaling, all the scaling approaches obviously mainnet rollups, let's say I'm a college student, I'm studying computer science, I'm fascinated by what's happening in Web3, I want to build decentralized Uber. If you were giving me advice, would you say go deploy that on a rollup? Go wait a few years? Deploy it on main? Deploy it across a few different rollups and mainnet. How would you think about that for the folks who are thinking around building applications for the mainstream?

Justin [00:23:49]

Right. So I would definitely encourage you to deploy on a rollup. And the reason is that very soon there's going to be a mass migration of users from L1 to L2. The reason being that general purpose rollups like Optimism and Arbitrum, which are live, and we now have incentives for people to move over to this new chain. So for example, the Optimism change just released the OP token and hopefully these two things combined will be sufficient to migrate network effects. And once you've migrated, there's really no reason to go back to L1 because you're just going to be paying 10 to 100 times higher transaction fees. There is this question around which flavor to deploy to - optimistic or ZK roll up. Well, today you have no choice. You have to use an optimistic rollup. But I think ultimately ZK rollups will win. But from the point of view of a developer, that's kind of a detail, right? Because a lot of these rollups are trying to target the EVM. So just build as if you're building on the EVM and then the infrastructure and the back end will improve over time.

So there's kind of two ways in which you can benefit from ZK rollups. Way number one would be for an optimistic rollup to upgrade to a ZK rollup. And I think this is the endgame for basically all optimistic rollups. Or maybe more plausible is that one of the ZK rollups manages to gain network effects. And the reason why I say more plausible is that you need to have a very, very strong technical team with very specialized knowledge in order to deploy one of these ZK rollups, so these dedicated teams will have a first mover advantage. So if you want to take a bit of a risk, then you can maybe try and bet on one of these ZK rollup teams, wait a few months until they're ready, and deploy there.

Katherine [00:26:00]

So now that we've covered a lot about The Merge, Ethereum, what it means move to proof of stake. I kind of wanted to switch gears and actually ask about the community behind Ethereum. So one of which is the Ethereum foundation, which is Justin where you're doing research. Can you tell us more about the Ethereum Foundation as an organization? So how it's involved with developing Ethereum or the Ethereum ecosystem? And just kind of tell us more about the foundation, about your role and how it really pushes technological innovation on Ethereum.

Justin [00:26:37]

So the Ethereum foundation is a not for profit, which is funded partly from the Bitcoin sale in 2014, I believe, and also from the adaptation of ETH that it received at the genesis of Ethereum. And basically its remit is to push forward the R&D of Ethereum, especially the R more than the D, the research. And also push forward public goods and fund them through grants. And so the foundation today has grown quite large. It has about 260 people. It's split up in many different teams that do different things. One of the teams, for example, is the grant giving team. And so here, basically what they do is process grant applications and try to find the best way to allocate a budget on the order of tens of millions of dollars every single year towards public goods that will support the ecosystem.

There's also various teams like the Solidity team and Geth team and the Socrates team, which is basically building core infrastructure for Ethereum consensus or peripheral infrastructure to support Ethereum. And then the team that I'm part of, which is the research team, which is, of size, roughly 20 people. What we do is try and use crypto economics, so basically both cryptography and incentives, to design systems that fulfill the criteria that we want around security and around scalability. And what we do is, the end product, is basically written specs. So we have these executable specs written in Python. And then the interesting thing is that these specs don't get implemented internally. It's a different foundation. Instead, what happens is that we have teams all around the world that do the heavy lifting of implementing these specs. It’s somewhat similar to the IETF, which is kind of a spec writing body for the internet. And then various teams like Google, you know, for Google Chrome, Firefox, Edge, all implementing these internet protocols in the context of a browser. And so one of the things that kind of comes out of this setup is client diversity. So we have, for example, five different teams that have been writing five different implementations in five different languages of the proof of stake spec.

And in addition to the research team, we also have a security team and a testing team. So one of the things that we found is that there are some benefits to centralization within the EF (Ethereum Foundation), specifically when it comes to security quality assurance, because it's difficult for every single implemented team to have security experts and to have testers. And so one of the things that we've been building in anticipation of the match is this security team internally to the EF that looks at all of the implementations in one go.

And so really, if you kind of zoom out, there's a bit of a pipeline, right? So, many years ago Vitalik planting the seeds of a vision, right? Even before Ethereum was launched, even before Genesis, Vitalik envisioned proof of stake and sharding. And then the goal of the research team is kind of to dream up these crypto economic games and then write them down and formalize them as a spec. And then that moves onto the implementer teams which are spread all across the world and have different sources of funding. And then it kind of comes back a little bit towards the EF for QA, audits, and things like that. And now when it comes to The Merge we’re at the very, very later stages basically of polishing and QA. All the heavy lifting has been done and in particular the research has been completely finalized. So one of the things you said at the beginning of the podcast is that it must be very stressful at the EF. Well for me I've done my work when it comes to The Merge, at least from the research standpoint.

Ash [00:31:19]

Now you're just memeing.

Justin [00:31:22]

Just memeing. Also helping coordination around QA and things like that.

Ash [00:31:28]

It's remarkable Vitalik seeing around corners, having a vision of the future everything you've done, the quality of folks at the EF. Just really incredible to see what you guys have done in such a huge lift in transitioning. There's so many doubters. It's never going to happen. This shift to proof of stake. So t's fun to see it's right around the corner.

Maybe just as a final question, and we talk about this a little bit in person a few weeks ago at Permissionless, but what is Ethereum in 10 years? In 50 years? How should we as participants think around what Ethereum will become or could become?

Justin [00:32:15]

Let's start with what Ethereum could be in ten years. So one of my theses is around shared security. This is the idea that there are network effects for applications being built upon a common platform, a common secure platform. If we have a blockchain which is simultaneously extremely high grade from a security standpoint and can scale to the whole world, then the network effects of such a platform would mean that it's kind of a winner take most. And so there is a possibility that in ten years, let's say, Ethereum becomes the substrate for securing the internet of value, basically.

Now, you may ask, hold on, the scalability roadmap that you put forward gives us 100,000 transactions per second. Is that really sufficient for the whole world? Now it turns out that if you allow yourself to project ten years into the future, there is another benefit. Scalability benefits basically another 100x. And the way that you achieve another 100x is by looking at the equivalent of Moore's Law for bandwidth. So that's a law called Nielsen’s Law. And it basically states that the consumer bandwidth grows 50% every single year. And this is something that we should expect to continue for at least a decade because bandwidth is this massively paralyzable thing. And even on this single strand of fiber, you can pass insane amounts of data - petabytes and petabytes and petabytes per second.

And it turns out that if you look at the fundamentals of scaling consensus, you need to look at the computational resources at hand, right? There's bandwidth, there's CPU, there's disc IO, which is basically reading and writing to disk, which is currently the bottleneck and there's storage. And three of those four computational resources can be removed as a bottleneck. Let me give you the example of CPU. How do you remove CPU as a bottleneck? The way that you do that is you use zero knowledge proofs. And so one of the things that we want to do on Ethereum is to basically snarkify the EVM, to create a ZK EVM in such a way that in order to become a validator, you only need to check these tiny proofs. You don't have to spend all the computation. And so that removes CPU as one of the bottlenecks for becoming a validator.

And then there's this other upgrade called statelessness, which removes two of our bottlenecks, namely the storage and the disc IO. And what are you left with? You're left with the fundamental computational constraint for scaling, which is bandwidth. And that, thanks to Nelisen’s law, is going to grow 100x over the next ten years. And so we're going to be in a position where we have 10 million transactions a second, which is more than enough for everyone on Earth.

I guess in ten years time, we can also make a prediction around the supply of ETH. So right now there's 120 million ETH and the supply might go down by let's say roughly 2 million ETH every single year. And so in ten years time, we should be around 100 million ETH in total circulation. And the fact that Ethereum kind of stands out as being the only chain right now that is basically capable of pulling such a feat of basically being able to extract more value in the form of the burn than value that it consumes just to secure itself and basically have this deflationary aspect to its supply may mean that ether the asset could become money for the internet. So basically we have this yin and yang where we have Ethereum, the platform, the blockchain, which secures the Internet of value. And then we have ether, the asset, which is optimized to be a collateral asset used in the context of staking by placing ETH as collateral to become a staker, but also used in the context of Defi, for example, in the case of decentralized stablecoins.

Now I guess the other part of the question was, what could be Ethereum in 50 years? And I guess there's kind of a couple of outcomes here. One is that it's become boring, right? It's become so pervasive and so commoditized within society that it's just this thing which is totally well understood. It's not changing. It’s totally immutable. And it's securing the internet of value. And that's kind of the happy case. The not so happy cases, that’s in 50 years is just completely forgotten and the relevance and there’s maybe a better piece of technology or some sort of competitor that takes over.

If we take the happy case that we can also look at what the supply of ether will be. What will happen over the long term is that the supply will find an equilibrium very similar to Bitcoin's 21 million limit. There will be an equilibrium for ether. And this is basically dictated by what percentage of the circulating supply is burned every single year. So if there's 1% of the circulating supply, which is burned every year, then the equilibrium is going to be 100 million ETH. If it’s 2% it’s going to be 50 million ETH. If it’s 3% is going to be 33.3 million ETH. My expectation is that in the endgame is that there's going to be  roughly 60 million ETH in circulation. And so just like Bitcoin kind of finds this equilibrium over decades towards 21 million through inflation, right? Bitcoin will be inflating for the next century, basically. It will be the very similar thing for ether the asset, but in the other direction. It will be deflating for 50 to 100 years and then eventually find this equilibrium, which could be, let's say, around 60 million. And so basically there's going to be a halvening, not of the issuance well of the supply, today we have 120 million ETH, and that could have to 60 million ETH over the very long term.

Ash [00:39:16]

And worst case, we end at 69 million ETH.

Justin [00:39:21]

Well, we do know that whatever happens, there will be some sort of equilibrium that that's the expected thing. So one of the things that Bitcoiners like to say is that the supply of ether is infinite, but if there is some kind of percentage of the circulating supply, which is burned every single year, then that gives us an equilibrium, whatever happens. And so that could be 200 million ETH, it could be higher than what it is today or it could be lower than what it is today.

Katherine [00:39:48]

Well, Justin, thank you so much for coming on today. We talked about a lot. We talked about The Merge. We talked about proof of stake. We talked about the future of Ethereum. And I think these are all really important topics to follow. Last question before you go - where can listeners go and follow the work that your team at the EF does or in general keep up with developments happening in the Ethereum ecosystem?

Justin [00:40:10]

Most of the researchers on Twitter, I'm @drakefjustin there. We also post on ETH research and in general, what tends to happen is that a lot of us work from home and then once in a while we go to a big conference. And one of the reasons why we go to a conference is not for the conference per say, it’s because immediately before or immediately after the conference we have some sort of research retreat. And so if you want to meet us a lot of us will be going, for example, to EthCC in Paris and a lot of us will be coming to DEFCON in Bogota. So, yeah, I'd love to meet you in person.

Katherine [00:40:51]

Awesome. Well, thank you so much, Justin. And we'll have you again. Some point again.

Justin [00:40:57]

Thanks for having me. Thanks, guys.

Katherine [00:41:01]

Thank you everyone for tuning in to episode four of Cross-Chain Examination. Please like and subscribe wherever you're getting this podcast, whether it's YouTube, Apple Podcasts, Spotify or the like. Leave me any thoughts, questions, suggestions for either this episode, past episodes or future episodes. Thank you again and I'll see you next week.

Expand to view full transcript
Collapse to smaller transcript view
accelerating the decentralized future
we strive towards the ideal. are you with us?