I just wrote my first story for MIT Tech Review.
It is an explainer piece on Ethereum’s move to proof of stake. What follows are notes from the story — along with additional ramblings and quotes from your favorite crypto skeptics.
When NFTs became a big thing in 2021, that drew a lot of attention to Ethereum, where most NFTs are traded. It also brought a lot of attention to the environmental horrors of proof of work.
Bitcoin and Ethereum both rely on proof of work to add new blocks to the chain. Together, they consume as much electricity as the entire country of Italy, according to Digiconomist.
Meanwhile, venture capitalists are shoveling cash at companies building Web3 — a supposedly new iteration of the internet where apps will run on permissionless blockchains, mainly Ethereum.
The problem is that permissionless blockchains — those that are open to the public and depend on a cryptocurrency to incentivize miners and maintain their security — are incredibly inefficient. They are sluggish. They can’t handle much data, and they don’t scale.
Case in point: CryptoKitties slowed the entire Ethereum network to a crawl in 2017.
In his article “The Web3 Fraud” Nicholas Weaver, a researcher at the International Computer Science Institute at Berkeley, explains that Web3 is “a technological edifice that is beyond useless as anyone who attempts to deploy a real application will quickly discover.”
Andreessen Horowitz (a16z), one of Silicon Valley’s top venture capital firms, is a big promoter of Web3. It has invested heavily in at least a dozen platforms that support NFTs alone, among them: Dapper Labs, OpenSea, Manifold, and soon, possibly, Bored Ape Yacht Club. Ethereum is crucial to a16z’s Web3 story.
Clearly, that story needs something more to support it. It needs a rocket-boosted ETH 2.0.
Scaling to the moon
In a proof of stake system, validators replace miners. Instead of investing in expensive ASIC systems that eventually end up in landfills, you invest in the native coins of the system.
Ethereum Foundation, the nonprofit behind Ethereum, says its proof of stake will consume 99.95% less electricity than proof of work. Ethereum currently handles roughly 15 transactions per second. Its founder Vitalik Buterin said ETH 2.0 could potentially handle a whopping 100,000 transactions per second. That would beat out Visa, which claims 65,000 transactions per second.
Ethereum was supposed to be a proof of stake blockchain from the start, according to its whitepaper. But in 2014, Buterin concluded that developing a proof of stake algorithm was non-trivial. So Ethereum settled for proof of work instead, while it went to work developing a proof of stake algorithm. Ethereum’s switch to proof of stake has been six months away for years.
Now, supposedly, the big moment is soon to arrive.
Ethereum is currently testing a proof of stake blockchain called the Beacon Chain. This will be the heart of ETH 2.0. So far, 9.7 million ETH ($25 billion) is staked on the Beacon Chain. To become a validator, you have to lock up 32 ETH. If you don’t have that much ETH on hand, you can join a staking pool.
After The Merge takes place, the next step is sharding — splitting the Ethereum chain up into 64 separate chains, so the network can scale. Sharding won’t happen until 2023. This is where the network reaches toward that theoretical number of 100,000 transactions per second.
Critics, however, doubt sharding will be any more efficient than a single chain.
Jorge Stolfi, a computer science professor at the State University of Campinas in Brazil, told me: “Almost every transaction will require updating two shards in an ‘atomic’ way (either both are updated or neither is updated). That will be the job of the central (Beacon) chain. I doubt very much that they can do that more efficiently than the current single-chain scheme.”
Ethereum, a centralized system
Scaling isn’t the only issue at hand in Ethereum’s move to proof of stake.
Proof of work’s decentralization suffers from economies of scale. Large mining operations are better able to maximize profits while lowering costs. This resulted in five mining operations controlling more than half of Bitcoin’s hash rate in 2020.
Like proof of work, proof of stake will naturally tend toward centralization.
Those who have the deepest pockets and stake the most coins will have the best chances of “winning the lottery,” thus reaping newly minted coins in the form of the block reward.
The big staking validators are already getting themselves into position. US crypto exchanges Coinbase and Kraken hold 78,000 out of 296,000 validators on the Beacon Chain.
A16z is also getting in on the action. It invested $70 million into staking provider Lido and is using Lido to stake an undisclosed portion of its venture arm’s ETH holdings on the Beacon Chain.
Proof of work and proof of stake both aim to get rid of a central gatekeeper, but that comes at a huge cost. One wastes electricity; the other wastes coins, which get locked up and pulled out of circulation.
“Whatever Sybil defense they use, economics forces successful permissionless blockchains to centralize; there is no justification for wasting resources in a doomed attempt at decentralization,” David Rosenthal said in a recent blog post. Rosenthal is known for co-creating Stanford University’s LOCKSS technology for the distributed preservation of digital content.
The one advantage of proof of stake that we can count on? At least it won’t destroy the planet.
If you like my work, please consider supporting my writing by subscribing to my Patreon account for as little as $5 a month — or more, if you are feeling generous!