“Blockchain” is something many have heard about, but complex to explain to non-technical people. Some time ago I decided to educate my family about how it works, to prepare them for the onrush of hyperbolic news they would hear about it in the following months. It took a few hours, and my explanations were followed by confused silence instead of the “eureka!” moments I was expecting. I had started that exercise with confidence, but the evening ended with much less understanding than I had planned for; all because of my dedication to technical details.
Focusing on the technicals is the worst way to explain what blockchain is and what it means. I am a programmer by trade, and despite that I have only a superficial interest in the inner workings of my laptop or phone – I just need to understand how they work on a high level to make use of them, and can leave the musings about circuitry to electronic engineers. Blockchain is the same: most people only need a high-level overview of its functional aspects because the technicals don’t really matter. No consumer of technology should need to know what a “nonce” is.
The purpose of this article is to offer an explanation of blockchain not by listing technical features, but through broad strokes, delving deeper only where useful.
Here is a summary for you:
- Blockchain is part of a family of technologies called “DLT”, or Distributed Ledger Technology. They are distributed record-keeping, a system in which multiple (potentially infinite) parties each hold a copy of the ledger.
- The “chain” part comes from the fact that blocks (i.e. pages in the ledger) are linked together cryptographically. If the ledger has 10 blocks, and you modify block 5, every block after that won’t check out anymore – it will be obvious that block 5 was changed to everybody looking.
- Transactions are approved by consensus: every would-be validator tries to be the first to solve a puzzle that gives them the right to choose what transactions will be in the next block (and collect the attached fees). Once a validator succeeds, the new block is broadcast to other participants in the network so that they can update their copy of the ledger, and after a while everybody agrees on the new state of the blockchain (a “consensus” has been reached).
- Blockchain is great when there is a lack of trust between parties that need a shared database to which they can all write (i.e. I have no reason to trust your version of the ledger because you could write anything in it, if everybody has a copy it’s easier to double-check). It’s massively redundant too, and can help getting rid of intermediaries and thus reducing costs.
- Blockchain is not-so-great when the issue is not a technological one, when decentralization wouldn’t help, and if intermediaries are not an issue. It is a tool that will solve some problems better than others, much like all other technology.
For those of you who have time, here’s a more detailed explanation.
What is “the blockchain”?
Although the two terms are often used interchangeably, “blockchain” is a specific technology that is part of a bigger family called DLT, which stands for Distributed Ledger Technology. The name highlights the most fundamental characteristic, from which its positives and negatives arise: it is distributed, i.e. participants in the network can host on their own machine a full copy of the ledger (“ledger” referring to a collection of records, most often accounting ones like transactions). For the most known blockchain, Bitcoin, any device can host a full copy of the its ledger, which contains a list of all Bitcoin transactions that ever happened.
In other words, DLT indicates distributed record-keeping, a system in which nobody holds the only copy in existence of the ledger, but instead multiple (potentially infinite) entities hold a full copy of it.
This has some interesting implications that we will examine in a few paragraphs. For the time being, here’s an analogy to understand the relationship between the terms used until now:
- If DLT is the concept of a plane (has wings, has engines, flies);
- a blockchain is a passenger plane (a type of plane with a specific structure and purpose);
- Bitcoin is a Boeing-747 (a specific type of passenger plane).
In this sense, both DLT and blockchain are theoretical definitions, whereas Bitcoin is one of the many practical implementations of those definitions. It is therefore generally useless to talk about just blockchain – it is much more important to talk about how the technology is used as a part of blockchain-based systems. In this series we focus on the more technical aspects, however, so we will mostly talk about blockchain the technology.
Why is it called blockchain?
The term refers to the structure of the ledger itself, where records are grouped together in “blocks”, cryptographically linked in a chain. Every block aside from the first one is linked to the block before and the block after through a pointer.
The process of linking the pages (“blocks”) in the ledger to one another varies, but I can give a simple example of a type of mechanism called “Proof of Work”. At the end of every page, the writer has to go through a few steps that produce a “summary” of its contents. Whoever writes the next page has to copy that summary at the top of their page, and produce a summary of their own page at the end, and so on.
- For each record (the 4 rows in the middle), write down the initial of the item, quantity, and total price. In the case of 5 kilos of potatoes at 1.25€, the result would be P,5,1.25;
- Sum all the results together, with the value of the letters being their position in the alphabet (so P = 16). For the potatoes, that would be 22.25, which then needs to be added to the other three;
- Add that sum to the “special letter” from the previous page summary. In our case it’s O = 15, so we get a total of 89.05.
- If the total can represent a letter (i.e.less than or equal to 26), that is the special letter of this page, and we add it at the end of the summary. Otherwise, we sum the digits of the total until we end up with a number equal or lower than 26, that can be transformed into a letter. In our case, 8 + 9 + 0 + 5 = 22 => V
- Repeat for each page. For the second page, the result is 22 (V) + 58.1 (records on the page) = 80.1. 80 is bigger than 26 so we sum the digits, and we get 8 + 0 + 1 = 9 => I, the special letter for the summary at the end.
Any change made to a page will change the total, which will change the summary and the special letter of that page. That, in turn, changes the copy of the summary in the next page, which (since the special letter has changed) also changes the total and special letter of that second page, and so on. Falsifying anything in the ledger is not a simple matter of swapping out a page in the middle, because the summaries keep them “linked” together so that they become invalid as soon as the page before is changed – anybody validating that the ledger is intact can spot the discrepancy immediately.
How does it work?
Recording transactions on a blockchain is fairly straightforward. As usual this can change between different systems, but there are usually only two parties involved, users and validators (often referred to as “miners”, but we won’t get into that here). Users are individuals/entities that perform a transaction; validators are those who hold a copy of the ledger, verify the validity of transactions, and record them on the ledger. Each validator has their own ledger, in which they add “draft” pages, which become “official” when accepted by a majority of validators. Note that there is no actual “official” ledger sanctioned by a central party: a page can be considered official when a majority of validators accepts it. This is called a “consensus”.
The mechanism by which consensus is reached is different from blockchain to blockchain. In Bitcoin, since everybody is trying to create a new block at the “open” end of the chain, at any particular moment the ledger looks more like a block-octopus than a block-chain. Block by block, this is resolved by reaching a consensus on which block everybody should be building upon, as there is an incentive in the system to collaborate and reach a common history. This is important because the shared history is what gives value to Bitcoin – without a central authority like in other systems, it is necessary to have some way to agree that you have three bitcoin and I have one, otherwise it’s impossible to do business.
Having a mechanism for agreement allows us to know that a particular transaction would be valid, and execute it. Thus the purpose of the blockchain is realized: a mechanism to share an agreement, a transaction in this case, without the need of a central power to ratify its validity.
A transaction, step-by-step
- I send money to you and we shout it out to all validators in earshot;
- Validators get a reward for being the first to add a new “approved” page in the ledger. They write the “summary” of the last page in their ledger on the new draft, and then fill it in by listing some of the currently unconfirmed transactions (i.e. not already on the ledger), which include ours. This is where the Proof of Work puzzle comes in;
- As soon as they’re done, they sign the page and send this draft, and a copy of any draft they receive and double-check, to as many validators as they can;
- The first draft to reach (and be double-checked by) the majority of validators counts as “official”. The validator who originally wrote the page is paid for their work;
- Because of this majority consensus, the new page is propagated until eventually all validators receive it and add it to their ledger. The transaction by which I sent you the money is now recorded in the blockchain. The process repeats.
What does it do better than other technologies?
Blockchain has many advantages, and they can all be attributed to its distributed nature.
No single point of failure, whether accidental or malicious
If a validator’s office burns down and the ledger is lost, there are copies in every other office. The ledger itself is very resilient to accidents because it has a huge number of “backups” all over the place. If a malicious actor breaks into an office because they don’t want a specific transaction recorded, they can’t simply “rip the page off”, because the page summaries won’t match (or won’t be correct), which will give the tampering away.
No need for trusted intermediaries
In the real world many transactions imply a degree of trust. For example in an online payment, the vendor trusts that the buyer has a certain amount of money in their account and therefore the payment is valid; this is guaranteed by the banking system, which is in this case the intermediary. As a bank won’t work for free, this trust comes at a price, and if you must go through multiple intermediaries, the fees naturally increase.
With DLTs, there is no chain of intermediaries to speak of – when your transaction is included in a page of the ledger that the majority accepts, it will eventually appear in all ledgers.
If you need to pay for a service or goods from Italy to Vietnam, there is a single step between buyer and vendor: the blockchain. Any cost that would come from needing other intermediaries is not required.
This is only the basic thrust of the argument, and there is already much discussion about other applications, for example disintermediating identification and land registries from governments, health data from healthcare providers, and more; one should however keep in mind that so far only monetary transactions have a proven track record of blockchain-based disintermediation, everything else is still at a very early or conceptual stage, and therefore unproven.
Transactions happen fast
The transactions in a block are generally considered “safe” (i.e. not reversible) after another 6 blocks are created on top of it. Even with a fairly long time to create a new block (for reference, the time to create a block in Bitcoin is around 10 minutes, for Ethereum it’s about 14 seconds) it takes about an hour for a transaction to be considered safe. As a practical application, transnational payments can be done on the blockchain at a fraction of the cost and time that a normal transfer takes.
Blockchain is not a silver bullet
As much as we believe in this technology and its many positive applications a number of misconceptions have arisen.
It’s not an appropriate solution for every problem
Despite its novelty and great potential, blockchain is just a tool, not a magic wand. It cannot (or rather, should not) be applied without a clear vision and analysis of the problem it’s trying to solve.
Blockchain can be applied as a new paradigm when some form of trust between parties that want or need to share a database cannot be guaranteed, or for optimizing existing processes, usually through disintermediation. These use-cases are fairly malleable in their scope (uPort for instance is building a solution for decentralized identity, po.et for proving content ownership, and there are many other examples), but blockchain is often inferior to other solutions that already exist. If all that is needed is a database with redundancies, blockchain is not the answer.
Another aspect to take into account is that not all problems are technological in nature: there is often a deeper, more human root to the issue, which will not be solved by better technology. For example, there are several projects that aim at giving identification to people who do not have it, like displaced persons, refugees, and so on. Lack of identification happens for many complex political, legal or social reasons, not for technological limitations: it’s not that governments lack paper, or computers, to provide identity to their citizens. Blockchain can address the technological part of the issue, even brilliantly, but it cannot provide an overall solution if the political, legal or social issues underlying are not solved as well.
Intermediaries aren’t evil
Blockchain is philosophically aligned towards decentralisation and disintermediation by design, and as a consequence intermediaries are often seen in a negative light, even though they are providing a useful function by filling a gap that exists in the execution of a process. Some of these gaps can be solved by technology (and the intermediary removed), as is the case with many payment solutions; others require a deeper shift than a simple technological upgrade, for instance improving the efficiency of utility companies. Having a centralized service is often the optimal solution to a problem, and allows for much more convenient interactions with the system.
Though information stored on the blockchain is not technically immutable, the more time passes the more difficult it is to change that information. This is a strength for the original purpose of Bitcoin (which wanted to offer irreversible transactions), but can be the source of many issues when the concept is expanded outside of day-to-day transactions. One day I might change my mind and want to delete a post I made on Faceblock; another, a purchase conducted on ethBay might be a scam, and I would rightly want my money refunded; the next one, a police officer might insert my ID on PoliceChain by mistake, and I now have a record without having done anything. One awful week, made possible by blockchain.
Jokes aside, solutions that allow to remove, scramble or hide data from blockchains are being explored. The concept is quite controversial, as the negative response to an Accenture paper detailing one such solution can attest. Taking that into account, the convenience of reversal afforded to us by centralized services might not be possible where blockchain systems are concerned, and the cost of putting countermeasures in place might offset the benefit from using blockchain in the first place.
Blockchain is a very promising technological innovation, with pros and cons like all human things, and it is shaping up to be a great tool that can be used ingeniously in humanitarian action and development. Explore the potential of blockchain for aid with us!