Breaking ChatGPT’s Bitcoin FUD #3: Scaling

According to ChatGPT, the limited number of transactions that the Bitcoin network can process might become a bottleneck as adoption grows. Which is a really valid criticism for Bitcoin, as no blockchain can scale. As a matter of fact, James A. Donald first said it to Satoshi Nakamoto on the cryptography mailing list back in November 2008: “it does not seem to scale to the required size”. This was the first public feedback that Satoshi Nakamoto received about Bitcoin and it concisely outlined the limitations of the system.

Interestingly, the same e-mail from James A. Donald also gave away a potential solution for wider adoption and scalability: “a file trading network akin to Bittorrent”. In other words, another layer that’s fast, direct in the way that it connects peers, and capable of handling a number of transactions which is only limited by every user’s bandwidth. Something that doesn’t need a blockchain and doesn’t become bloated and centralized if there’s too much activity going on. Also, a system that’s very much like Bittorent in its design and therefore doesn’t make every user store, verify and validate everyone else’s data. This all sounds an awful lot like the Lightning network, doesn’t it?

James A. Donald’s technical criticism was truly visionary, as he predicted both the scaling wars that would take place in Bitcoin’s first decade of existence and he also suggested a way to increase the velocity of the currency. However, this 2008 statement alone doesn’t break ChatGPT’s FUD – a proper contextualization is required in order to highlight the current strategy for scaling Bitcoin and why it’s very likely to work.

First of all, why does Bitcoin deal with scalability issues at all? It’s all in the nature of decentralization and how blockchains work. The first instinct is to make the network mine larger blocks at a faster rate – but doing so only makes it more difficult for regular users around the world to validate their own transactions like every sovereign bitcoiner should. Conversely, the other side effect of increasing the block size is the requirement for more powerful hardware – with a slippery slope which ends at servers and industrial-grade computers which effectively defeat the purpose of decentralization. Once you get to a certain block size (as BSV did), you might as well use PayPal.

The ideal network needs to be secure, decentralized, and scalable. But according to Zooko’s popular triangle, computer networks can only have two of them and must choose their tradeoffs – with the intention to complement them in some other way. By far, security is most important. Without it, financial transactions have no finality and are therefore worthless. But physical cash and gold also have good finality. So it’s decentralization that represents Bitcoin’s main value proposition – it accomplishes to offer access to a global financial network without the need for banking institutions and with no controlling third party actor in the middle. Anyone can participate in mining and validation, anyone can download a wallet and receive coins even if they delegate validation to another node (SPV, or Simple Payment Verification). Nobody can censor a transaction, confiscate the funds, or restrict specific individuals from using the network. Without these qualities, Bitcoin is no different from PayPal.

But all of these come at the expense of limited scalability. Bitcoin mines a new block at an approximate rate of 10 minutes and adds about 2000 new transactions that receive confirmation. The system is not terrible for this early stage of adoption. Its design with an arbitrarily limited block size is supposed to enable a burgeoning and competitive fee market which supports the miners’ revenue. However, Bitcoin must be able to accommodate all financial transactions of the world in an uncompromising way – and it accomplishes this feat through layers.

In time, using the base layer will become too expensive for most people around the world. Thankfully, everyone will have a Lightning channel or sidechain to use with various tradeoffs: for the sake of speed and cost-efficiency, these Bitcoin layers can be more centralized. Since security must remain a constant and the highest level of decentralization is also the most expensive way to transact, it will be the latter value that gets lowerred in order to increase scalability and convenience.

Though lowering the decentralization is not ideal and might become dangerous in some instances, what matters the most is that users get to choose between competing layers and they also benefit from the option to not rely on a trusted third party. For example, users can already opt for custodial Lightning solutions – which offer cheap, instant, and private transactions at the expense of trusting the custodian. What keeps entities such as Wallet of Satoshi, Alby, Tippin and BottlePay honest is the competition from other similar and non-custodial services, as well as the looming threat that users will open their own Lightning channels and therefore stop supporting the service with their transaction fees.

The same goes for sidechains, federations, Chaumian mints, and other parallel chains. Since it’s in their best economic interest to have a large user base and stay competitive, they are far less likely to exit scam. Of course, this doesn’t sound ideal – but it’s better than Bitcoin’s current dominant scalability option which involves centralized over-regulated exchanges that act as banks and alternative blockchain networks which make use of experimental features and onboard newbies in a way that provides education with both using wallets and getting scammed.

Increasing the blocksize to fit more transactions is not completely off the table in the distant future, but no reasonable size can accommodate all financial transactions around the world. When storage gets even cheaper and 5G internet connections become common with affordable data plans even in the least developed nations, the community might decide to hard fork in order to increase the block size and therefore allow more users to open Lightning channels and peg into sidechains at lower costs. For example, 32 or 64 megabyte blocks might be reasonable in 2040. But if on-chain transactions keep on getting smaller and more efficient via soft forks (as was the case with SegWit in 2017), we might not need any kind of hard fork.

The biggest issue with increasing the block size is not storage or bandwidth, but social consensus and the slippery slope theory. Some will argue that making such a significant change will only prove that it’s possible to coordinate to arbitrarily change Bitcoin – and therefore generate more such similar events to the point where the block size either gets unreasonably big or else the community gets too fragmented while supporting different versions of the chain. It’s much easier to deal with parallel chains and second layers, whose potential failures do not break the base layer’s security and decentralization.

Furthermore, increasing the block size does not guarantee that the blocks will get filled all the time. Such an action directly generates lower transaction fees for miners due to lower competition to get into the next block. During times of low demand for block space, when the coinbase rewards will already have been cut in half a few times, this can effectively pose a security threat to Bitcoin.

To directly address ChatGPT’s FUD, Bitcoin is scaling and will get better at scaling as more layers get developed. The limited number of transactions that can fit into a block can be Lightning channel openings and sidechain peg-ins which endow users with the ability to transact in a fast, affordable, and scalable way. The only issue is that these layers come with various decentralization tradeoffs. But they’re still more decentralized than altcoins, centralized exchanges, and other forms of trying to poorly replicate or else replace Bitcoin with poorly designed systems. The criticism is valid, but it lacks context and nuance. So once again, ChatGPT takes an L.