The principle of trias politica, or separation of powers, refers to the division of government into three branches so that each arm can check the powers of the others. A legislative power creates and changes laws. A judiciary power interprets and rules on laws. An executive power administrates and enforces the law. Together, they enable greater a distribution of responsibility across a wide range of organizations and individuals.
Previously, I touched on PoW vs. PoS blockchains and blockchain scaling solutions. Today, I will elaborate on how I think the blockchain industry might develop to make an impact on business operations and consumer experiences in the future, layer by layer. I’m an optimist when it comes to blockchain and cryptocurrency, so I invite different opinions on everything I touch on below!
Layer 0, The Executors
In the crypto industry, Layer 0 is often referred to as the physical devices and connections needed to establish a blockchain. In other words, computers and cables (and internet). In my commentary on proof-of-work (PoW) blockchains, I discussed why I believe hardware is critical to network security. Using Bitcoin as a model, I pointed out how physical limitations mitigate certain forms of attack, and all but nullify others. In my view, the establishing of a hardware layer is necessary for an expansive, healthy, and secure blockchain ecosystem.
But why would anyone invest in the hardware to support a blockchain? Especially when a lot of the skepticism surrounding cryptocurrency today is centered around the question of value - why does ‘X’ token have any value?
When the world wide web was first gaining traction, there was similar doubt surrounding a lot of the new internet-first companies entering the market. I recall friends and family reminiscing about how the likes of Google or Amazon weren’t considered “investable” in the late 1990s/early 2000s. Most didn’t understand their revenue models, and those who did probably didn’t find them more appealing than tried-and-true names like Exxon or Johnson and Johnson (at least those companies make products that people use!) In reality, they weren’t wrong at the time - Amazon had its first profitable year 7-years after IPO - but a prescient mind back then could piece together just how digital our lives were to become. A healthy dose of foresight and optimism is required to make bets on the distant future.
I fully understand the skepticism surrounding the “fair value” for some blockchains and tokens. It’s not immediately clear whether or not this technology has any staying power, let alone the ability to make an impact on any other industry. However, as uncompelling as it may seem, a lot of the value many investors, technologists, and enthusiasts see in blockchain today is future value. Just as Amazon and Google (and Rome) weren’t built in a day, new applications for blockchain are being implemented all the time. Some will stick and most will not, but I would argue that blockchain is here to stay due to its potential to revolutionize digital security and uniqueness. Security is an obvious need, but what do I mean about uniqueness?
Layer 1, The Arbiter
Blockchain is an incredible technology that enables a game theory-protected form of digital scarcity and uniqueness. Why do these features matter? When you stream a movie on Netflix, you don’t care if your TV sources the original movie file or a copy of it on a server somewhere. But when you send your bank a message to complete a wire transfer, you certainly hope that those details aren’t being copied to a different server somewhere else. The same might be true for the deed of trust on your house, your medical records, and your social security information. You want those items to be secured from bad actors that might use them for malicious purposes.
While nearly all of your private information is being kept on a server somewhere in the world, it’s (ideally) being kept securely by a party you trust - like your bank, your doctor, or your government. But those trusted parties are either getting worse at securing data, or bad actors are using more sophisticated methods to illegally access data (probably both). Even some of our must “trusted” institutions are vulnerable, like the US Government, NATO, and Microsoft. Certainly there must be a better way to retrieve to data than disinterested employees and bad passwords?
Blockchains might not be the best way to store large amounts of data, but they are a compelling format to secure access to important data. Passwords are an outdated method of securing data, because there are so many ways to crack them. It’s why major tech companies are moving towards passwordless sign-in technologies. Blockchains are somewhat unique in that every user is given two “passwords” for each account - one that you show everyone, a public key, and one that you share with nobody, a private key (experts will know I’m oversimplifying this, but bear with me). With those two keys combined, a user can submit a transaction or signature to the blockchain to confirm their identity. Further, every single transaction and signature is recorded on the blockchain with perfect accuracy, so you can trace every single action made by an account (i.e. no more “that wasn’t me, I was hacked” excuses).
The way many password systems today work is essentially as follows: You tell Company X your password when you set up an account, so that you can log back into your account and they know it’s you (obviously). The benefit of this process is that Company X can help you recover and change your password if you forget it or lose it. The issue is that now your password is kept on their database and lives within their security environment - whether you trust it or not. As a hacker, I would much rather spend my time trying to crack a database full of passwords than to attempt to break into an individual account.
By securing access to private data using blockchain, the benefits over legacy systems are twofold: Security and transparency. When it comes to most public blockchains, each user is in total control of their account (i.e. their public and private keys), making it incredibly difficult and unprofitable for bad actors to crack a single account. Of course, if a user tells someone their private key, then that account is forever compromised - or if they lose it, then the account is lost. Admittedly, giving users total sovereignty comes at a cost, but I think it is worth the risk. And with the improvements being made to biometric logins across all our devices, I think managing public and private keys will quickly become much more accessible to all users, regardless of technical ability.
As it pertains to transparency, because every transaction and signature is stored on the blockchain forever, it would be a simple task to trace every data request and authorization. While your bank information and medical records might still be stored on a server somewhere, accessing that data might require your digital signature on the blockchain. Since only you control your private key to authorize access, investigating stolen data becomes much more traceable - questions like how/when a hacker accessed the data are answered by records on the blockchain.
For this reason, I think a layer 1 blockchain should focus on two issues in the “blockchain trilemma” - security and decentralization. If these two items are the top priority, then scalability and functionality can be addressed with other solutions.
Layer 2, The Creators
Of course, any software is only as valuable as its application - this is where scaling solutions fit in. A blockchain that focuses on security and decentralization at the expense of scalability must have some way to add features that make use of those qualities. A layer 2 expands the capabilities of its host with additional applications and ways for users to engage, and doesn’t necessarily need to focus too much effort on security or decentralization. This is where I think proof-of-stake (PoS) is best suited.
A network of PoS layer 2s that are hosted by a secure PoW layer 1 seems like an optimal setup. By using a different consensus mechanism than layer 1, these layer 2 networks can facilitate faster transactions for “semi-finality” on their own chain, that are then confirmed on the slower (but more secure) PoW layer 1 host network for permanent finality.
I specifically call for a network of PoS layer 2s on the host blockchain, for two main reasons:
Competition is important for innovation
Development of a market for block space on the layer 1
The first point is semi-obvious, yet why do I advocate for a single layer 1 network to rule them all? Because the layer 1 is more infrastructure than application. Plumbing and electrical infrastructure are specifically tailored to each individual property, but some sort of genericness in the parts used is critical to building things in a cost-effective manner. Of course, for more specific use cases, a separate layer 1 might be necessary - and that’s fine too! But for the most part, a layer 1 that hosts an ecosystem of layer 2s is probably all that’s necessary (for now).
The second point is more nuanced. First, what is block space? Well, put simply, it’s the available space for data on each block that is added to the blockchain. Just as your iPhone can’t hold thousands of high quality images due to limited storage space, a block cannot hold every single transaction waiting to be processed. By limiting the amount of space a block can hold, a transaction fee market develops. Those who want their transaction processed faster must pay a higher fee to be included in the next block. With a multitude of layer 2s competing for block space to achieve permanent finality for their applications, there can exist a more robust market for block space. If only one layer 2 existed, they would likely not have to pay much to be included in each block, since they would only be competing against simple transactions on the layer 1.
A developed block space market adds to the security and decentralization benefits provided by the layer 1 network, since more miners would be incentivized to go into business if they feel that the transaction fees being paid outweigh their operating costs. More miners improves the security offered by the host blockchain, which would ideally attract more layer 2 networks to build on it. Over time, the amount of incoming miners would equalize with the amount of growth the layer 1 and layer 2 networks experience.
These layer 2s would also provide smart contracts, scalable applications, and more features to its layer 1 host. Another benefit to hosting an ecosystem of layer 2s rather than a single layer 2 is that applications can be siloed to a specific blockchain. Provided the layer 2s have a way to communicate some information to each other (like ownership of certain assets), there isn’t much benefit to having all kinds of applications on a single blockchain. In fact, having a layer 2 blockchain developed for a specific purpose is likely to yield better results than modifying an application’s functionality to fit an existing blockchain’s capabilities.
For example, there might be a layer 2 network focused on a borrowing and lending protocol, and another that only serves a metaverse gaming project. While they are both hosted on the same layer 1 for the security benefits, the prerequisites to users joining the layer 2s might be different - like some form of ID for the financial protocol and a specific set of NFTs for the metaverse. These two blockchains would compete for finality on their layer 1 host by paying more or less for block space, despite not competing with one another for clientele. This helps develop a block space market for the layer 1 host blockchain, while not having much impact on the end users of each of the layer 2 services.
Because the layer 2 networks have different objectives for achieving finality on the layer 1, they would each have different fee tolerances when it comes to securing block space. A financial protocol might pay higher fees to attain finality on a regular basis, whereas a metaverse gaming project might not be so pressed to achieve that same consistency. As a result, the fee revenue generated by those projects (from their users) could be more dynamic based upon the fees they pay to be included in blocks on the layer 1 blockchain. This would then develop a series of companies that differentiate themselves by quality of service. A top of the line financial protocol may charge higher fees for their service than a lower end competitor, because they pay more for guaranteed block space to achieve finality on their layer 1. (Just like how a Porsche is in many ways superior to a Toyota, but that added quality comes at a cost.)
Vertical vs. Lateral Separation
A question I continue to think about is why a vertically layered approach might work better than lateral approach. In other words, why would a network of layer 2 blockchains on top of a layer 1 host be better than just a bunch of independent blockchains? If I argue for an ecosystem with a single point of failure (i.e. the layer 1 blockchain), wouldn’t some sort of standardized communication between individual blockchains be better?
I keep an open mind on this, but still err towards vertical layers for two main reasons:
Enforcing software communication standards across a wide ecosystem of decentralized teams is easier said than done
A lot of the industry currently uses the lateral approach, and we continue to see hacks of bridges between blockchains (i.e. the communication protocols that connect separate blockchains)
I think a vertically layered approach forces the establishment of some sort of standardization for layer 2 protocols - after all, they need to communicate with the same host. Further, since the ecosystem is built on top of the same layer 1, a lot of the back and forth between layer 2s might be done on the host blockchain - thus eliminating the need for bridges in many cases. Already, hundreds of millions (if not billions) of dollars have been stolen by attacks on blockchain bridges. They make for obvious targets for hackers, as their holdings are visible on two different networks, and many bridges still operate with some human involvement (i.e. they are not totally automated and are therefore prone to social attacks).
Of course, if some revolutionary blockchain intercommunication technology comes along, my opinion might change. But for now, a vertically layered approach mitigates some of my concerns above and ties the PoS layer 2 economies to tangible inputs. The expense of powering the PoW layer 1 network must, in some way, affect and drive costs for the layer 2s, contextualizing digital economies in the physical world.
For now, I think separating the “responsibilities” of each layer makes the most sense moving forward. Just as trias politica reduces the concentration of power in government to good effect, I feel that the same principle applies to blockchains. End users generally don’t care about the infrastructure used to deliver their products. In other words, most people don’t worry about whether or not their internet browser is delivering content from the nearest server, and I don’t imagine most blockchain users will care if the layer 1 network is fast or highly customizable - as long as the applications they use on layer 2s are. For this reason, outsourcing security and decentralization to a layer 1 affords greater flexibility to layer 2s to develop more engaging and useful products and services.
Foundations
Of course, my concept of trias reticula (or “three networks”) is simplified here, since my aim for this piece is to articulate a general framework for the future. This is a much more intricate proposition, and I admit that I have glossed over some key points while completely skipping others. There is far too much for me to condense into a single article.
Some might find my views simplistic. Others might find them complicated. I plan to elaborate on this vision in the future, but hopefully I’ve successfully communicated a strong enough foundation to build upon. Any questions or comments you might have after reading this are essential for me to continue the thought - please share them!
For the blockchain enthusiasts reading this, where do you feel I’m oversimplifying or overcomplicating the blockchain “layer cake?”
For the blockchain skeptics, what other systems are better suited to handling some of the problems I’ve described?
For the totally uninitiated that don’t understand any of the above, please let me know what I can explain better! I want to make this as accessible as possible, because I honestly believe that this technology might make for a safer, more efficient future.
And to everyone who made it this far, thank you for reading!