Today’s public blockchains are called “world computers”. The most famous one is Bitcoin, the first cryptocurrency based on a distributed public ledger.
The unique thing about Bitcoin is that it was the first cryptocurrency to solve the double-spend problem in a trustless, decentralized manner. Bitcoin itself took the blockchain and used it for a single purpose: A peer to peer, decentralized electronic cash. That’s it.
The second most famous blockchain is arguably Ethereum.
Ethereum employs a blockchain just like Bitcoin, but aspires to do much more: Create a general purpose, decentralized world computer.
To achieve this, they have built a robust infrastructure that is capable of running complex operations in the form of smart contracts. The two “killer-app” examples for Ethereum’s functionality are ICOs (a 6.8$b a year market in 2017!) and Cryptokitties (Trading virtual assets. in this case, cute cats).
Ethereum is an amazing feat of engineering — a global world computer that works through a vast ecosystem of smart contracts.
But there’s a caveat.
General Purpose vs Specialized
In the world of computer hardware there are roughly two types of hardware: General purpose and specialized.
A great example of “general purpose” hardware is our commonplace computer CPUs. These chips are at the heart of any computer and smartphone and are flexible enough to handle any kind of work thrown at it.
The benefit of general purpose systems is that it enables creators to solve most tasks using just software — the CPU handles any manner of tasks you throw at it.
The drawback: The CPU is a jack of all trades but a master of none. If you are looking at performance intensive tasks such as graphics rendering or cryptocurrency mining, CPUs are slow and inefficient.
On the other side of the spectrum you have “specialized” hardware. A good example of this category is GPUs.
GPUs are designed for rendering graphics and are the best hardware in the world for doing just that.
The benefit of specialized hardware is that you can tailor your solution to a specific set of problems, benefitting from significant performance advantages at a lower operational cost than general purpose solutions.
The drawback? Specialized hardware is only good at the work it was built to do.
We can see this taken to the extreme with Bitcoin miners — ASIC chips specifically designed and tuned to mine Bitcoin, and nothing else.
The Problem with General Purpose Blockchains
The truth is that today’s blockchains don’t scale well yet. We see that with Bitcoin, and we see that 10x with Ethereum.
The problem is that proof-of-work blockchains are not good at scaling up, especially if you are aiming to be a general purpose world computer — a blockchain Swiss Army Knife.
Just to illustrate, today, the Ethereum public network is insanely inefficient when compared to traditional cloud computing (Amazon AWS):
Cost per hour of a medium Amazon AWS instance: $0.04
Cost per hour of the Ethereum world computer: $41.6k
As you can see, we are talking about a 1,040,000x overhead.
But what if the right solution to scaling blockchains is not having a single be-all-end-all chain, but creating specialized chains for each specific use-case? Instead of scaling one blockchain to the extreme, why don’t we just add more, as needed?
In essence, this is the market that we have been building for the past few years — we have hundreds of separate projects tackling various blockchain uses-cases today, but it’s general purpose platforms that get all the attention.
It’s hard to say what the right solution is, but it sounds like there is a lot of potential in treating each use-case for the blockchain separately, and achieving a truly decentralized “world computer” by combining hundreds of different blockchains, each one built for a different niche.
Going with this approach, you automatically get less traffic per blockchain than you would if we used Ethereum for everything. Also, this approach opens up technological pluralism: Each project is free to approach its’ own technology and scalability issues in any way they choose, depending on said project’s unique requirements.
For example, what if we create a separate blockchain just for tokens and crowdsales? Provided it was decentralized and had enough mining power behind it, such a solution would inherently be more efficient at tokens and crowdsales than Ethereum is. And more importantly, a simpler blockchain is easier to secure because it only does a few things. That’s what makes Bitcoin itself so elegant — it’s built to do just one thing: Transacting digital money.
Pluralism is Good
Having a pluralistic approach towards the frameworks that power the decentralized economy is a good way of maintaining true decentralization.
Today, if we embrace Ethereum, we embrace Vitalik. But if we embrace blockchain “pluralism”, we embrace the economy at large.
This is not to say that there is no place for Ethereum and general-purpose blockchains. On the contrary, Ethereum and other platforms can be great solutions for many use-cases, especially projects that don’t need to deal with bootstrapping a blockchain. But there should be a limit on how much trust can be put in one place.
Before the market can truly support many blockchains, we need to crack blockchain interoperability, and that’s exactly what we are working on at Safebit. Come check us out 🙂
Are General Purpose Blockchains the Right Solution? was originally published in Hacker Noon on Medium, where people are continuing the conversation by highlighting and responding to this story.