How to actually solve cryptocurrency’s scaling problem
It won't happen overnight. The real cutting edge is the research that lays the groundwork for the future.
Scalability is cryptocurrency's biggest problem, but creating a network that's sufficiently decentralised and practical isn't easy.
There's no single right way to solve the scaling problem, although there are definitely some wrong ways, and different projects are approaching it from different angles.
Consensus mechanisms are the heart of the scaling problem. Consensus mechanisms are how all the different nodes of a decentralised network can "agree" on what's happening and work together with complete strangers all over the world to ensure the network is actually decentralised.
Each has its own pros and cons, and different ways of hopefully finding a solution to the scaling problem, and includes elements of engineering and economic theory to secure the network.
- The engineering takes the form of meticulous programming to ensure that a network does exactly what it's intended to do, without any exploitable loopholes.
- The economic theory elements are there to hopefully incentivise people to do exactly what the network intends them to do, without letting them exploit any loopholes.
The first and best-known consensus mechanism, used by bitcoin and many others, is called proof of work.
Proof of work (PoW)
This type of system uses cryptocurrency "miners" to uncover blocks and add those blocks to the blockchain. This is done by solving math problems with computing power. The first computer to find the right answer will get to add the next block to the blockchain, winning a bitcoin reward for doing so. These blocks carry transactions across the bitcoin network.
On paper, this is highly decentralised because theoretically anyone can be a bitcoin miner. But in practice, only a limited number of people will have access to enough computing power to make it worthwhile.
At the same time, the economic theory behind PoW is dependent on the mined cryptocurrency remaining valuable enough to motivate miners. This means it has to remain scarce no matter how much energy is poured into the network. It does this by dynamically increasing or reducing the difficulty of the math problems based on how many machines are trying to solve them.
The end result is an inherently inefficient system. PoW consumes a lot of energy, but all that effort is dedicated to solving those math problems to mine blocks, rather than directly process transactions.
Bitcoin is looking at solving its scaling problem through a kind of secondary off-blockchain network called the Lightning Network, but the inefficiencies and limitations of proof of work have seen many other cryptocurrencies explore other solutions and write off proof of work as an evolutionary dead end that won't work in the long run.
As such, a lot of the research in the space right now is focused on alternative consensus mechanisms, which broadly refers to anything other than proof of work.
Proof of stake (PoS)
Proof of stake (PoS) is the most popular alternative consensus mechanism and is widely considered to be one of the most promising avenues for the near future.
This mechanism removes the concept of cryptocurrency mining and replaces it with "staking."
This is when people put proof-of-stake cryptocurrency into a specially designed PC wallet, and then connect that wallet to the network where it automatically becomes a node and starts using that person's computing power to process transactions along the network. For doing so, users are rewarded by earning interest on their stake.
Just like PoW miners, stakers are rewarded with cryptocurrency for their services. The upside is that unlike PoW miners, they're using their computer's power to directly process transactions rather than just solve math problems, making it potentially much more efficient and scalable. At the same time, requiring users to stake funds opens up a greater range of behavioural engineering possibilities, such as penalising potential attackers by taking their money.
The main downside, other than the potential centralisation problems of the "rich get richer" model, is that it leans heavily on potentially unpredictable economic theory and manipulating people's behaviour to keep the network secure. Functionally, proof of stake is only as good as its ability to program human behaviour with economic incentives. This leaves it potentially vulnerable to unpredictable elements like people who are willing to spend millions or billions of dollars to destroy the network or otherwise act "irrationally".
If there's one advantage to proof of work, it's that it has been tried and tested and definitely can be theoretically secure and decentralised.
Proof of stake doesn't have the same distinction, says Aparna Krishnan, a Thiel Fellow, co-founder of Mechanism Labs, recipient of a UC Berkeley Blockchain Lab Grant, founder of the education department at Blockchain at Berkeley and all-around expert in alternative consensus mechanisms who's currently focused on proof of stake.
Progress and challenges in PoS
There are many PoS cryptocurrencies in existence, Krishnan says, but none of them have been stress tested to the same extent as PoW.
"There are functional proof-of-stake protocols that aren't necessarily secure, but they are proof of stake," she said. "NEO calls itself proof of stake, but I wouldn't consider that secure at all... There has been a huge rise of other PoS consensus tokens... a lot of them have been working silently behind the scenes. I would not regard them as secure, but they exist."
And even though PoS is her current focus, Krishnan doesn't necessarily think it's going to be "the one".
One of the more interesting problems, she says, is that some of the inherent vulnerabilities of proof of work are also confounding research in proof of stake. PoS is built around the idea of humans as economically rational machines, but by the same logic, most PoW cryptocurrencies should be smouldering train wrecks by now, which they simply aren't.
"One of the most interesting things I've come across is that there are a lot of theoretical attacks on bitcoin or blockchains in general that don't seem to happen in practice. I don't know if we just can't detect them happening, or if they're not happening. It could be either," Krishnan says. "Which means humans are not economically rational, or maybe it's just too expensive [in terms of money, time, expertise or anything else]."
Krishnan also cautions against the idea of being too focused on purely economic incentives in PoS.
"When people think incentives, they just think economic incentives. I think we can have others," she said. "Incentives are incentives, and there are multiple kinds of incentives. I don't think it's fair to say economic incentives are the only kind."
The psychological elements involved in securing networks add a whole new dimension to an already-complex area, so a lot of research, rather than just building, is going to be needed to find an effective solution.
"There's a lot of open research happening, but I think people underestimate the rigour you need to build such a system," Krishnan says. "For example, if we're going to be having entire economies on the blockchain, I don't think intuition is sufficient rigour to build them out. There's going to need to be a lot more academic and security analysis of these systems before we build them out."
This is one of the areas Mechanism Labs is currently focused on – a research-oriented exploration of alternative consensus mechanisms as a way of solving cryptocurrency's scaling problem.
It might also be more likely to make a dent in the problem than other research groups. True to cryptocurrency form, Mechanism Labs is the world's first open-source blockchain research lab with research papers being collaboratively written through a GitHub repository.
Measure twice, cut once
Mechanism Labs' open-source studies are more than a spiritual nod to cryptocurrency's usual open-source workflow. It's designed to help bridge the gap between industry and academia through transparency, while softening the learning curve for those who want to enter the space and quickly getting them up to speed on the current cutting edge, rather than just re-treading water in proof of work. With a lack of experienced blockchain developers significantly slowing research and development, these kinds of initiatives are critical for the development of the space as a whole.
"What we're primarily focused on [at Mechanism Labs] is scalability, but our research focus is alternative consensus mechanisms. But Mechanism Labs as a whole is supporting all kinds of research that is focused on scalability," Krishnan says. "Mechanism Labs is the first open-source blockchain research lab. What we're doing with Mechanism Labs is we're bridging the gap between academia and where industry is at today. There's a huge lack of rigour, also there's a steep learning curve for anyone who wants to get involved in the research."
One of Krishnan's current projects, Meta-Analysis of Alternative Consensus Protocols (PDF), is similarly focused on doing the hard yards and laying the research groundwork for quicker future developments in the space through a comprehensive meta-analysis of alternative consensus protocols.
Just like writing things down meant humanity didn't have to rediscover metallurgy, aluminium and flight before inventing spacecraft, it means new researchers to the space won't necessarily have to rediscover proof of work in order to discover proof of stake.
This kind of research-centric approach is especially useful right now, Krishan notes, given the differences in how different countries are approaching blockchain and cryptocurrency. Some countries are jumping in with both feet, while others are taking a wait-and-see approach to see how the technology shapes up, to let some complete solutions emerge and see what it can actually be used for.
As national attitudes towards blockchain and cryptocurrency continue crystallising, research itself can still be valuable and used to accelerate development and future research everywhere.
"I would say [views on blockchain] drastically varies from country to country," Krishnan said. "When I went to teach executive education in India... they're quite aware of all the developments happening, but they're quite sceptical of building on blockchain technology until its completely ready.
"In terms of what's happening in the Middle East, that's a very different scene. Dubai is very much hyped about blockchain as a technology, they're trying to build use cases as quickly as possible. [But] when I went to do a couple of presentations at banks in Oman, they wanted other people to push out before they even start using it. But what I think every country can do more is push forward innovation, so that will speed up the process of having a blockchain platform ready for all the use cases a country wants."
Diversity from the ground level
Cognisant of the need to lay groundwork for development of the space as a whole, Krishnan has also taken a keen interest in the current gender disparity of the blockchain research space. It's a matter of practicality as much as anything else, with a huge body of research showing that improved diversity across almost any measure is typically being associated with preferable decision-making outcomes.
Blockchain technology, as an almost brand new field of research, presents a unique opportunity to try to pragmatically bake in a level of diversity from the beginning.
"It's only when the field is starting that you can make the huge changes and bring in more of this diversity, so I'm definitely quite supportive of more women in the space," Krishnan said. "It's not just going to happen one day, it requires deliberate effort [and] a field in its inception where it's possible to make these kinds of changes."
The first step, Krishnan says, is to notice it.
"I think earlier on when I started education at Blockchain at Berkeley, it never occurred to me that we don't have sufficient women. It just wasn't a thought that came to mind. But there are definitely times that you realise that another woman's perspective would be valuable on a lot of issues.
"I didn't realise the huge gap initially. But as soon as I realised it, I started mentoring more women in the field. At any point in time, I'm always mentoring four to five other women in the space... doesn't have to be the same area, just enabling them, seeing how they're doing and connecting them to the right people, the right researchers. I do also encourage other people that I mentor to keep the cycle of mentorship going."
Laying the groundwork and ensuring a more diverse field of researchers and developers from the start might have longstanding benefits for the space as a whole. And with the widespread adoption phase still a few years off by Krishnan's and other experts' estimations, there's never been a better time.
Finding the killer apps
There's also the question of finding blockchain's "killer apps," and actively hunting down and testing them as soon as possible. These killer apps will be the Googles, Facebooks and Ubers of blockchain technology, potentially changing the world in many very real ways with their systems.
Pretty much any cryptocurrency project that isn't an outright scam is currently trying to do exactly this, but it's difficult with the underlying technology still in flux, and a more comprehensive body of research is likely to see more use cases emerge.
"I think blockchain as a technology can be ready in a few years if we push," Krishnan said. "I definitely think research on concepts is very useful... also testing out use cases and figuring out what the killer use case is for blockchains, and then building out blockchains related to that in parallel.
"I think we haven't found all the killer uses for blockchain technology yet. It's like when the Internet started. Did people imagine Uber and Facebook?
"I think currency is one of the immediate use cases. Right now, the big use is going to be payments. I'd say there will be a lot of changes in supply chain and in general there's going to be a huge new mechanism of crowdfunding [that] will change the way we do research, will change the way maybe educational institutions exist. Maybe even [change] government structures. But I think all of that is way way down the line.
"We're still in the hype stage for a while. Even though it appears it's slowing down, we’re still very much in the hype stage. I think the hype should not affect the technology builders though; they shouldn't be there for the hype, they should be there for the future."
Disclosure: At the time of writing, the author holds ETH, IOTA, ICX, VET, XLM, BTC and ADA.