Beyond Bitcoin: The Blockchain Trilemma

Chapter 13:

Previous: How to buy Bitcoin on Binance with DCA auto-invest

The theory goes that an Internet of Blockchains will eventually emerge as the dominant crypto infrastructure. An interconnected, diverse system with dozens, maybe hundreds of different ecosystems with different features, pros, and cons. In many ways, as we will see in later chapters, it is already happening, and it is my preferred outcome. After all, if decentralization is the objective, why not have a decentralized web of decentralized projects? But there is a problem.

In this chapter, we will specifically address the problem faced by the Internet of Blockchains. Vitalik Buterin, the co-founder of Ethereum, gave the problem a name. It is called the Blockchain Trilemma. To go beyond Bitcoin and to understand more crypto projects better as we discuss them, we need to learn about this issue.

Furthermore, in the two chapters that follow this one, we will learn about two more vital aspects common to almost every blockchain project that came after Bitcoin. These are proof of stake (PoS) consensus, the main alternative to proof of work, which we have already learned about, and smart contracts, which turn these modern PoS networks into global decentralized computers.

Back to the problem.

Contents

The Blockchain Trilemma

The Blockchain Trilemma refers to the three-pronged features we associate with blockchains and crypto networks. We have actually already covered some issues when talking about the Bitcoin block-size wars.

An excellent way to start the explanation would be with an image.

The blockchain trilemma

The image shows three aspects of blockchain and crypto that we have discussed. Decentralization, security, and decentralization. These attributes are correlated, complimentary, and connected. For example, the more nodes on the Bitcoin blockchain, the more decentralized. Simple. However, when we want to scale a blockchain for more users, the trilemma rears its ugly head. As you move towards scalability, at the top of the triangle, at least one of either decentralization or security must suffer.

As an example, you might remember that with Bitcoin, the argument was about increasing the block size and reducing the block time to increase the number of potential transactions. The idea for scaling up was simple, boosted transactions. The argument against it was more nuanced. If the block size and time are “upgraded,” the hardware required to run a node becomes more expensive, fewer people run nodes, and decentralization is sacrificed.

Layer 2 solutions

As we saw in Chapter 5, the solution, or at least the most currently viable solution, was the Lightning network, built on top of Bitcoin.

As we will soon explore, other blockchains can also have networks on top of them. Take Ethereum, the number 1 decentralized computer on planet Earth and the second biggest crypto project by market cap. Ethereum is way faster than Bitcoin, with a new block every 12 seconds! But even that is not fast enough if everyone starts using DeFi and NFT games. Ethereum couldn’t even cope with an explosion in growth of the Crypto Kitties NFT project in 2018. Ethereum gas prices for exchanging tokens regularly reached hundreds of dollars per transaction in the 2021 DeFi peak.

Ethereum is adopting a raft of layer two solutions. There are more technically correct ways of describing these varying scaling solutions, and we will dig into the details in later chapters. They include sidechains like Polygon, which processes transactions and hosts dApps, then batch reconciles transactions onto Ethereum. Other solutions include advanced cryptography like ZK Rollups, allowing for off-chain processing and super-fast on-chain reconciliation. These solutions add capacity off of the Ethereum main chain but guarantee the final transactions on-chain. In addition, Ethereum is pursuing advanced solutions like sharding.

Sharding

Ethereum is one of many projects looking into sharding. Near Protocol is another project promising vast TPS or transactions per second increases by splitting each block into multiple pieces called chunks and verifying them all separately before recombining them all. At the time of writing, Ethereum is at least a year away from completing the sharding upgrade, and Near has sharding working but with only a minimal number of shards. The problem is allowing all the shards to communicate with each other. What if we could just have a super-fast layer 1 and forget about the complexities of layer 2s and sharding?

Super-fast transactions

Solana is a layer one blockchain that claims to be able to do 65,000 transactions per second. This sounds highly scalable. However, the problem with Solana is centralization, at least relative to Bitcoin. As you move up the trilemma triangle towards scalability, something has to give.

First, the node software is developed only by Solana’s developers. This means there is one version of the node software. In 2022 I lost count of how many times I read the news that the Solana network was “down,” and a centrally coordinated rebooting of the software was required to get it up and running again. Add to this that the hardware requirements to run a Solana node are enormous, including 12 core – 2.8GHZ CPU, 128 GB of RAM, and the storage requirements were so advanced I didn’t even understand them. Add to this that these requirements will likely increase as the network grows. This means there will be many times fewer nodes than Bitcoin, which can run on a Raspberry Pi.

What else might help the world of crypto onboard the rest of the planet?

Secure multi-party computation

Secure multi-party computation is not new and it is slow. It requires the nodes of a network to constantly confer with each other, and worst of all, to share data, likely private data. The newest solution I have seen, although unfinished and unproven, is NMC or Nil Message Compute. The theory is that with NMC technology, you can divide the computation steps into many tiny parts, guaranteeing that a bad actor cannot manipulate the transaction. NMC guarantees accurate processing of each tiny piece, nullifying the need for the nodes to reach consensus, at least for some computations. The recipient of the result can quickly verify that the result is genuine. This means that distributed consensus is unnecessary because it is mathematically guaranteed, and trust is transferred to complex unbreakable math. The most likely use of NMC technology is as a computation layer that a blockchain could optionally use to lighten the load or for particularly high-volume use cases. Another potential use is to store our private keys and seed phrases. It is hard to predict what the final infrastructure will look like. This is one reason why investment is so precarious. I would happily bet on crypto, but deciding which projects to bet on is much more challenging.

Interchain and multichain

Another potential solution is to design layer one from the start to interoperate with multiple other layer 1s. Examples of this are Cosmos and Polkadot. Both these projects are discussed soon. Either or both of these projects could be the blockchain network of the future, but neither is there yet.

The point is, however, that despite mass innovation and incredible emerging solutions, if planet Earth decided to adopt one blockchain tomorrow, or even if the Earth’s billions, formed orderly groups and spread themselves evenly between the existing solutions; the whole thing would grind to a halt! This is because none of them are ready.

I didn’t write this to disrespect Solana, Ethereum, Polkadot, Cosmos, Near, or anyone; likely, all these projects could play a big part in the future of crypto. Instead, the point of telling you about the problem is to understand why there is such a proliferation of layer 2s, roll-ups, and innovative layer 1s.

Summary

We have learned about the Blockchain trilemma and the tradeoff between security, decentralization, and scaling a network for billions of users and that the future of the technology is uncertain. Soon, we will explore projects specifically that chose different tradeoffs as solutions to this trilemma and some which even claim to solve it on layer 1.

As we know, Bitcoin achieves consensus between the nodes on the network by Proof of Work mining. However, most other cryptos use a different method called Proof of Stake. So let’s talk about it.