Ethereum is slowly moving to proof-of-stake. Is that good news for climate?
You just have to doom-scroll through your Twitter feed to realise that the last few weeks have been a turbulent period in cryptocurrency. Non-fungible tokens or NFTs, units of crypto purportedly acting as simulacra of digital objects ranging from digital artworks to in-video-game weapons, to tweets, to farts, have exploded in notoriety, seeping well into the mainstream. (You can hardly get more mainstream than John Cleese, or Christie’s.) Yet, digital artists who have been making a killing by piggybacking on the nifty craze have reportedly been grappling with a creeping feeling of guilt: their new way of monetising their work is wasting energy and damaging the climate.
That comes hardly as a surprise to anyone familiar with how cryptocurrencies work. The first and most popular cryptocurrency, Bitcoin, is designed as a peer-to-peer payment system that does not rely on any single arbiter – like a bank or a financial intermediary – to validate payments between users. Instead it is underpinned by a decentralised swarm of computers collectively maintaining a historical log of payments – the Bitcoin blockchain – and updating it periodically via a process akin to voting to approve new transactions.
One potential pitfall with that model is the “51% attack” scenario, in which a malicious actor deploys a battalion of sock puppets to take over a majority of the network and vote through transactions that should not be approved, for instance because a user does not own the bitcoin they want to spend. To stave off that risk, Bitcoin is designed to make participation in the network expensive: computers that upkeep the ledger, also called mining nodes, are required to constantly try to crack daunting mathematical puzzles, receiving Bitcoin rewards upon solving them. In so doing, these (fairly expensive) computers consume a lot of electricity, a set-up intended to render attempts to tamper with the Bitcoin ledger too onerous and encourage cooperation instead. But this electricity-based incentive system, called “proof-of-work”, means that today Bitcoin mining consumes 133.65 terawatt-hours a year, more than the annual consumption of countries like Sweden or Ukraine, according to the University of Cambridge’s Centre for Alternative Finance (CCAF). Only 39 per cent of that electricity comes from renewable sources, a 2020 report by the University of Cambridge found.
According to Michel Rauchs, a CCAF research affiliate, Bitcoin’s success story set proof-of-work as the go-to cryptocurrency “industry standard”. Ethereum, the second most traded cryptocurrency, and the platform where a large chunk of NFTs are being minted and traded, also adopted proof-of-work when it launched in 2015 – hence crypto-artists’ remorses. By one account, an NFT sale on Ethereum can result in the consumption of as much as 8.7 megawatt-hours of electricity, more than twice what an average British household consumes in one year.
Yet, that might be changing. Ethereum is in the midst of an overhaul aimed at replacing proof-of-work with a less energy-guzzling security make-up. If successful, the move could add further momentum to an ongoing shift away from proof-of-work, and reduce crypto’s carbon footprint and wastefulness.
This mechanism is called “proof-of-stake”. Where proof-of-work is underpinned by real-world hardware and energy expenditure, proof-of-stake grounds its effectiveness in the perceived value of cryptocurrency – in this case Ethereum’s ether – and in game-theoretical trade-offs. Mining nodes are replaced by validators, required to put down a substantial sum – currently, 32 ether, equivalent to over £38,000 – as a deposit to participate in the network. The deposit, namely the stake, accrues interest over time, thus providing an incentive for validators to join. The race to solve mathematical problems is replaced by an algorithm appointing validators at random to wave through transactions, which are then confirmed if approved by two thirds of the network. A validator attempting to approve irregular transactions or otherwise acting dishonestly gets a fine – paid out of its stake – and in the most serious cases loses all of its stake, and is cut loose from the network.
Ben Edgington, a project lead at ConsenSys – a software developing company involved in the roll-out of Ethereum’s Eth.2.0 upgrade, which will introduce proof-of-stake alongside other major changes – says that while a 51% attack is still conceivable under proof-of-stake, it would require attackers to stake twice as much ether as the rest of the network combined. “So if there’s $10 billion (£7.3bn) worth of stake securing the chain, I would need $20 billion (£14.5bn) in order to be able to do any meaningful attack on the chain,” he says. An additional assumption is that the price of ether would plummet when an attack was reported, so pulverising the value of the attacker’s stake. Like in proof-of-work, incentives have been stacked against cheating – but this time there are no thrumming, electricity-hungry mining computers involved: one can theoretically set up a validator from a laptop.
The main disadvantage of proof-of-stake is that it is new and complicated, says Danny Ryan, a researcher at the Ethereum Foundation. “Proof-of-work is a lot simpler: the complexity of proof-of-work gets offloaded into this physical hardware and the actual software design is much simpler. One of the biggest criticisms I can make of proof-of-stake is that it’s a much more complex design,” he says.
That not only means that it is harder to convince people of the soundness of the network’s design, and that there is no handy mining metaphor to deploy, but also that the system is less well-tested than proof-of-work, and some ways of taking down a proof-of-stake system haven’t been discovered yet. “So far, proof-of-work has been around for 12 years now. And we know that it works,” says Matt McKibbin, founder of tech advisory firm DecentraNet and a proponent of using cleaner energy for proof-of-work mining as opposed to doing away with it.
Indeed, getting proof-of-stake right has not been easy. At Ethereum’s inception in 2015, its creators had promised that the system would switch from mining to staking fairly quickly. That, Rauchs says, might have played a role in the choice of new cryptocurrency projects, then mushrooming at an impressive pace, to opt for a proof-of-stake mechanism. Several high-profile blockchains, indeed, such as EOS or Cardano, use various flavours of proof-of-stake. “Ethereum’s implicit assumption that they will move towards proof-of-stake in the future, really set up an alternative view in motion,” he says. “Increasingly new developers and companies that were building their own blockchain networks started to immediately ditch proof-of-work from the initial design.”
Yet, Ethereum’s deadline for shifting to proof-of-stake kept being postponed. Ryan says that the reason was, bluntly, that designing a proof-of-stake blockchain while preserving security and decentralisation – two of cryptocurrency’s seminal principles – was daunting. “When ethereum launched, there was no version of a secure proof-of-stake,” he says. “There were proof-of-stake algorithms, but there were very like known theoretical critical issues and attack vectors.” Given Ethereum’s quick rise to the status of the world’s second blockchain, Ryan says, the developers involved had to tread carefully.
Now, however, it looks like the shift is getting closer to becoming a reality. In December 2020, Ethereum launched a blockchain powered by proof-of-stake, which in time will become the backbone of the entire network as proof-of-work is phased out. How long the transfer of the entire Ethereum ecosystem to the new make-up will take is unclear, but Edgington thinks it will be a matter of “months, not years.”
“Basically 90 per cent of the work is done,” Edgington says. “We just need to move over the existing blockchain on proof-of-stake, and the job is done.” Rauchs, the Cambridge researcher, says that he finds that timetable “optimistic, given the complexities associated with that.”
The direct carbon footprint effect of Ethereum’s shift will be sizable, but still just scratching the surface of cryptocurrency’s energy consumption problem – for which Bitcoin is still mostly to blame. Rauchs says that it is estimated that as of last year Ethereum consumed between 20 and 25 terawatt hours annually; that is roughly one sixth of Bitcoin’s expenditure. “That will be two months of Bitcoin operating. So I don’t think it’s that significant in that regard – that we’ll drastically reduce carbon emissions, simply for the reason that Bitcoin accounts for pretty much the vast majority of it,” Rauchs says. “But, from a symbolic point of view, I think it could definitely help moving away from proof-of-work in the future.” This shift – when it happens – might send ripples through the cryptocurrency industry, making proof-of-work rarer and rarer.
The big question is, of course, whether Bitcoin will ever take the leap to proof-of-stake itself. That sounds incredibly improbable, given the place proof-of-work and its intuitive, hardware-backed architecture holds in the Bitcoin community at large. But as Bitcoin starts gaining ground as a new asset class among institutional investors and technology billionaires including Elon Musk, it is not totally unlikely that something will have to give. For instance, Rauchs says, can Bitcoin be considered an ethical investment, or will it fall foul of so-called ESG (Environment, Society and Governance) criteria?
Matt McKibbin, founder of tech advisory firm DecentraNet and a proponent of green cryptocurrency mining, thinks that, in a way, a transition to a system minimising Bitcoin’s carbon footprint is inevitable. In fact, it was the point all along: McKibbin argues that Bitcoin’s huge energy footprint will end up accelerating the switch to renewable energy. In 2018, in a Medium post entitled “Satoshi’s Second Gift: The Renewable Energy Bounty Program”, McKibbin laid out his theory on how Bitcoin’s pseudonymous inventor Satoshi Nakamoto, inadvertently or deliberately, encouraged clean energy production. From his perspective, Bitcoin rewards are acting as a lure towards using cleaner – and generally cheaper – renewable energy sources such as solar or hydroelectric power.
“There’s a point at which Bitcoin becomes a ‘bounty programme’ for the cleanest types of energies in the world,” McKibbin says. “[That’s] because it makes it economically viable to figure out exactly how to do those energy systems in the best possible way, because they’re cheaper than fossil fuels.”
