Is Amazon single-handedly propping up blockchains?
April 7th, 2022
Welcome back to The Delegate! Previous recipients of this email have seen a line that says we’re in “experimental mode”. I’m here to make good on that promise. We tried the weekly digest, and people said they wanted deeper dives on specific topics. So let’s try it.
Upcoming topics: Polygon and the future of Ethereum scaling; Application-specific chains; Pocket Network and decentralizing Infura; A vision for Play-to-Earn; Copying is a good thing, actually
If you want to see something in particular, hit reply and let me know. And if you dis/like this new format, let me know using the links at the bottom of this email.
In November of 2013, Vitalik Buterin published the Ethereum whitepaper, titled "A Next-Generation Smart Contract and Decentralized Application Platform", the foundation for our understanding of web3 today.
8 years later, in 2021, a spike in crypto interest rocketed costs to use Ethereum to untenable heights. Cargo cults touting their preferred Ethereum "killer" picked fights with each other.
One common line of criticism: “your chain isn’t decentralized”.
But what makes a chain decentralized or not? What will it take to realize Vitalik's original vision of a decentralized application platform?
And how close are we to getting there?
A decentralized future
It's helpful to start by imagining what a decentralized blockchain could look like in 30 years.
It’s 2050. More than half of the world interacts with web3 on a daily basis. The most used blockchain processes immensely more scale than we can imagine today. It relies on the work of hundreds of millions of nodes that participate.
Many, but not all, nodes are actively making changes to the blockchain’s state. The rest simply follow along, copying changes when they're confirmed.
In 2022, the people getting this email are mostly only users of the blockchain. Some of us don’t have any crypto at all, others have used a few apps. None of us are actually part of the blockchain.
But in 2050, all of us will be actively maintaining the blockchain itself. The technology will be so easy to run that our phones themselves will be nodes making changes to blockchain state.
That's what we're headed.
Web3 use cases aside (we know it still has a lot to prove), the technology has to be simple enough that everyone can run a node. And the economics have to incentivize people to want to participate.
How are we doing so far?
If today were 2050, the most used blockchain would be Ethereum, and we’d hope that Ethereum has hundreds of millions of nodes, some or most of them our mobile phones which are actively maintaining the blockchain.
Will we have our hopes dashed?
In short: If today really were 2050, we would be miles behind. Or in web3 terms, "we're still early".
Though the exact number of active Ethereum nodes can be hard to decipher, current dashboards show the number around 3,000-5,000.
Around 70% of active Ethereum nodes live on cloud hosting providers.
Plus: Half of cloud-hosted nodes are on Amazon Web Services.
Meaning: We're short on nodes by 1,000x. And, ironically enough for a movement touting decentralization from day one, if AWS disappeared tomorrow, it would take a chunk of Ethereum nodes with it.
That's not quite as decentralized as we imagined it.
By the way, the Ethereum "killers" are even worse off.
(In fact: Solana's scaling strategy is to rely on higher and higher-end hardware over time. Today, the baseline requirements include a 12-core CPU and 128GB of RAM; a custom PC with these specs would start at $5,000)
Does that mean crypto has failed?
I wouldn’t say so.
My take: Calling crypto "fully decentralized" or debating which chain is more decentralized than another both feel premature.
There are real concerns about the state of decentralization today, be they related to the number of nodes or the fact that a lot of them live on AWS.
But writing off crypto because it's not there yet also feels premature.
The visions painted by web3 are titanic in size. There's still years of technical work before the movement has a real chance at realizing those visions.
I'm optimistic: Crypto developers have already shown promising steps that the decentralized world we imagine is possible.
Mina, a new blockchain project, keeps the entire size of the chain to 22KB (smaller than this webpage; and vs. 6TB for Ethereum), making it possible for phones to be nodes
Vitalik’s publicly stated position for Ethereum is that all designs need to make it more possible for the average person to run a node.
So we might see a lot more movement on this front in the next 2-3 years.
Like scalability, decentralization is an existential point to get right, yet it poses immense challenges that will take years to solve. I’m eager to learn about those efforts as they develop. We’ll check in in another year or two for updates.
What’d you think of this week’s update?