2:2 Public Ledgers Matter
t some point 30 seconds or less into a conversation about DLT this question will inevitably arise:
So, I love this whole distributed ledger idea, but do we need “cryptocurrency” to implement a DLT solution?
To understand the shape of our fitness landscape, we must tackle the deeper issues this seemingly innocuous question poses.
Phrased less innocuously, we can read this question as:
Will the value creation accrue on “public” open source ledgers?
OR will valuation creation accrue on privately controlled ledgers?
And of course we can’t forget the most subversive hidden meaning:
Do we need the blessing of incumbents to adopt decentralized infrastructure, or will new natively decentralized businesses emerge to take their place?
Now having read half of this book, you calmly start the conversation by clarifying when we talk about “cryptocurrency” we really mean:
a shared global ledger that anyone can write to as long as they abide by the rules of the network.
currency is the most simple use case, but by no means the only..
intellectual property fingerprints.. health records.. real estate titles, turing complete virtual machines that can execute any arbitrarily complex contract…
While an open source paradigm was the first practical implementation of DLT with the Bitcoin protocol, there is nothing to stop the technology from taking a decidedly private turn.
Many in the open source community default to an open-is-always-better mentality. While a noble ideal, such thinking limits deconstructing why private networks exist, and the potential benefits of accruing value behind private walled gardens.
Before breaking free of the technical entirely to have fun in the socio-cultural, we should run down how private DLT networks are created, and what features set them apart from their public counterparts.
The main difference between public and private networks generally comes down to the notion of “access tokens” which give certain users elevated rights on the network.
Like their public counterparts, as transactions are hashed together, no individual user can tamper with data on the ledger (modify or delete privileges), as their copy of the ledger would not match with other copies on the private network.
The key difference is read and write access can be given or revoked at any time by administrators by issuing or revoking access using token mechanisms. If you have an active token you can read, write, or both. Without the right access token credentials, you can find yourself suddenly out of the network with no recourse to recover data on the network.
In traditional corporate power structures, it’s obvious why such a system architecture is preferred.
In a supply chain context, each member in the consortium of suppliers would receive credentials to use the shared ledger. Otherwise, anyone from the public could in theory post transactions to shared supply chain which could cause havok in the system. If a supplier lost a contract renewal, rights could be easily revoked to prevent them from posting/reading subsequent transactions on the network.
In a banking context, access tokens get more controversial as the core tenant of open ledgers like bitcoin is to give anyone in the world peer-to-peer access to banking services that cannot be revoked by any third party. In a permissioned banking ledger, accounts can be frozen, or funds stored on a users wallet transferred without the users permission depending on the level of rights given to the administrating bank.
Regardless of your feelings on the philosophical promise of Bitcoin being co-opted for corporate use, the closed nature of permissioned ledgers provides use case flexibility that open ledger do not have. There are inherent security drawbacks in purely private implementations however.
Namely the nodes that validate transactions are by their very nature incentivized to co-opt the ledger. Think about the core value add of any distributed ledger network:
be it an inter-bank consortium for real time settlements
to an inter-supplier consortium to share royalty information about oil and gas flowing through under sea pipelines.
The whole point of the ledger network is to serve as neutral Switzerland that disparate parties trust to store critical shared information. Hypothetically then, what happens if:
one party (or a few parties in collusion with each other) create their own version of the truth
then force the network to agree with them?
As we discussed in part I, on the bitcoin network this requires a sustained attack to build the longest main chain that becomes the primary source of truth. While the winning chain will reject transactions from the losing chain, the loser can still fork off into a new ledger that maintains an unadulterated record of past.
Private ledger networks typically use different consensus mechanisms where a very high percent of nodes are required to make consensus, or else the entire network grinds to a halt. In these systems, history can be written over when other nodes on the network replace their older version of history with an updated version.
In a pure private network with only a few partners running nodes, it is not out of the question that an altered version of the ledger can become the dominant ledger if there are enough incentives for parties on the network to collude with each other.
The solution to this problem is quite simple - write way point transactions (hashes) to public ledgers!
As alluded to several times in Part I of the book, there is a marriage of public and private ledger networks where each system can do what it’s best at. Private ledgers can scale, while public ledgers can serve as a court system that guarantees the authenticity of private ledgers.
This is accomplished by taking a fingerprint of the private consortium chain at regular intervals and storing that hash onto one or more public networks.
For instance every batch of 10,000 transactions might include a special fingerprint hash that is inserted both into the private network, then into a public network. If at any point in the future shenanigans occur, anyone running a simple verification program can determine if all of the hashes match. Such a system precludes any history modifying attacks as it would be nearly impossible to take over one public ledger network, let alone several.
This is technically possible due to a critical mathematics component of distributed ledger networks we discussed briefly in part I - the Merkle tree. Merkle trees allow you to aggregate fingerprints (hashes) together in a binary tree of transactions, where only the Merkle Root (or genesis of the tree) is needed to prove the entire branching network of transactions.
By placing Merkle roots (super compressed fingerprints) into public blockchains, we can efficiently prove to the world that private transactions are immutably valid.
So the real answer for where value creation will occur is BOTH on private networks, and on public networks. As public networks by their very nature are open to tragedy of the commons issues, incentives must be in place to to prioritize important data like fingerprint hashes, over flooding the shared commons with uncompressed cat pictures.
Unlike centralized storage solutions where you pay a monthly or yearly subscription fee to store data that can be deleted if the renewal is not paid, permission-less decentralized data storage is a pay-once/store-forever model. E.g. if you pay the mining fee once to submit a transaction, your data will be stored on say the Bitcoin blockchain forever. Moreover, if your data ever differs from the other ledgers by even a single byte, that node will not agree with the others and its transactions will be considered invalid. To prevent spam data like malicious uncompressed cat pictures from eating up too much storage capacity on each redundant full node, the Bitcoin blockchain keeps the mining fee per byte very high to discourage placing large volumes of data onto the main chain.
Conversely, on a permissioned ledger, data CAN be modified or deleted with enough consensus from the ruling administrators. Even if the ledger is append only, a new chain can be voted in to replace data on the previous chain making the hashed links between each transaction on the previous chain obsolete. (That is unless fingerprints were stored on enough surviving public ledgers)
With a fingerprint internet, we can easily end up in a two tiered system where:
Closed source (proprietary) logic runs on private ledger networks
Open source logic runs on a public ledger networks
If this sounds familiar, this is exactly how the current internet works.
Companies like Google with a proprietary PageRank Algorithm and Ad network use many open protocols such as TCP/IP, Apache, etc.
In this system, close to 100% of the value created is in the private network layer, with next to zero value creation in the public layer as there was no inherent way to monetize the shared commons.
The code that comprises TCP/IP for instance is a publicly funded research effort where companies make donations to the Internet Engineering Task Force which implements upgrades to the protocol. Such a system is imperfect, as like in politics donations can sway the development of the internet to favor certain companies over others.
Thus to understand where the most value will accrue in the new ledger internet, we must have some reasoning behind why people would want to use private ledgers over public ledgers, or vice versa.
This reasoning can be summed it up in a trite platitude.. Open Source = Trust
Imagine you have two competing alternatives to store something really important like a real estate title, or college degree.
The first system stores your title using opaque blackbox logic on its own permissioned private chain, and shows you a hashed receipt securely stored on a public ledger.
The second system stores your title using open source logic on its own permissionless side chain, and also shows you a hashed receipt securely stored on a public ledger.
The first systems claims it has a proprietary compression algorithm that makes storing and retrieving title/college degree information much more cost effective than the competition, but of course cannot reveal exactly how it works.
Meanwhile, the second system is publicly auditable by anyone in the world to ensure it works as advertised (and does not have some terrible concurrency bug in it that allows a hacker to steal the title to your house, or pretend she has your college degree)
Of all the paradigm shifts brought in by the DLT revolution, this notion is often the hardest to swallow. If the goal of DLT is to unlock value by replacing fallible black boxes with transparent trust systems, public ledgers should accrue the majority of the value over the long term. Alas, ledgers are ultimately social technologies, and there is no predetermined default where the more egalitarian system wins. Even though Google PageRank is a proprietary blackbox, people still trust them to organize the world’s information.
A Tale of Two Business Models
To illustrate this point, let’s examine two competing real estate title registries (or college registrar registries.. its all the same thing) each with a competing philosophy on closed vs open source. Both projects have put much time, effort, and investor dollars towards:
creating friendly user interfaces
developing marketing plans, then paying exorbitant fees to attend crypto conferences
working with individual county clerk offices/colleges to integrate with their legacy systems.
The open project: would most likely issue a “utility token” that allows users to place real estate titles/college transcripts into their ledger. Each time something is placed into the ledger, a micro fee is generated that gives value to the token. This system by design is open to anyone to add entries to, which then requires a micro fee to prevent the system from being spammed with fake entries. While the data is publicly available for anyone to read, it is encrypted thus meaningless unless the owner of the entry provides the decryption keys.
The closed project: meanwhile would probably pursue a more traditional Software-as-a-Service model and try to win consulting & subscription fiat dollars for on boarding new clients. In this system, spam is less of an issue as only approved users with access tokens can use the ledger. Unlike the wild west of the open system, if there is a mistake on the ledger, administrators can rollback changes, or revoke access, depending how the system is designed.
Depending on the use case, one type of system might gain more traction in the marketplace over the other. College degrees might lean open, while real estate titles might lean closed.
Both systems automate the middleman by providing a quasi-neutral to truly neutral shared commons to transfer data. which unlocks vast value creation.
In the case of the real estate title, an automated trust system could eventually replace the 1000+ dollar per transaction title insurance industry with a vastly cheaper and more effective solution.
In the college registrar use case, the back offices of the college administrations become vastly more streamlined, reducing transcript fees towards 0.
But what stops another project from coming in and copy pasting the open source code that powers the open source project, then marketing their clone better? Doesn’t the closed source project with a proprietary competitive moat around their product have an advantage?
Survival of the Trusted
Capitalists typically think of value creation as the sustained competitive moat using something proprietary like special logic, or branding to keep their margins from eroding. In Zero to One, Peter Thiel famously brought the term “unfair competitive advantage” into the popular lexicon.
At the end of the day, what is the difference between Bitcoin and an identical copy of Bitcoin? From a source code standpoint nothing, but from a social value and branding standpoint everything.
Not-Bitcoin might only have a few dozen miners around the world supporting its security, and might only be traded on a few exchanges, while Bitcoin has billions of dollars in daily liquidity and burns as such electricity as a small Scandinavian country.
Which platform do you want your critical fingerprints stored on that secures your real estate title? Bitcoin or Not-Bitcoin?
That is the strongest argument for betting on open source public protocols, over closed sourced private networks. Open Source = Trust Us beats out Closed Source = Trust me. Why?
There is no barrier to entry to join and grow the trust network.
Open systems are constantly attacked and hardened, while closed systems atrophy in the darkness.
While this simple logic helps prove the value of public protocols in general, it does not prove the value of any specific public protocol. What people trust today is not a guarantee of what people will trust tomorrow. A perfect example is the war brewing between the dueling visions of Bitcoin. One favors off-chain scaling and a small main chain, while the other favors scaling up the current system by processing more transactions per block. Over time the market will decide which strategy wins, or will decide neither electricity heavy strategy makes sense and switch to less energy intensive systems.
With the ever present threat of a copy/paste with tweaked parameters to lower fees towards zero, we are left being pulled both towards:
a perfectly competitive commodity market where lowest cost per transaction is king
or monopolistic competition where brand matters and trust is worth something more than the lowest cost approach.
If it was your real estate title or your college degree, what type of system would you want protecting your legacy?
Critical to the success of one project over another, is an incentive structure where it is more profitable for speculators, developers, users, regulators, incumbents, and other market participants to join a project, rather than try to build their own:
open/closed source version
The computer science principals we learned in Part I have no bearing on which of these philosophies is used (E.g. functional programming principals can be deployed into a closed source - private - permissioned system just as easily as a public - permissionless - open system) These issues are ultimately more socio-cultural, than purely technical as it would seem on the surface.
In the next chapters, we will dive further into how trust is created structurally inside of distributed ledger projects by analyzing the various ways projects are funded, governed, and propagated into the market. This analysis will hopefully guide us towards heuristics for what winners in the DLT fitness landscape might look like.