Changing Bitcoin, Part II — The Present

“It used to be very hard to change Bitcoin. It’s still very hard to change Bitcoin, but it used to be very hard to change Bitcoin too.”

– paraphrasing a joke of comedian Mitch Hedberg

One of the most profound differences between Bitcoin and nearly every other human invention is how hard it is to make changes to it. Since Bitcoin is software, this aspect of its nature makes it profoundly different from other software. Software, after all, is constantly changing. Software is expected to get upgraded regularly, even automatically. Software is accepted as being imperfect and in need of regular patches. And, software is understood to be something that will become obsolete, eventually being replaced by some other software.

None of that is true in the case of Bitcoin. Bitcoin rarely changes. It is extremely hard to change.

But if software is easy to change, and Bitcoin is software, there must be something else about it that is unlike all the software that came before it, that makes it an exception to this rule.

Indeed, there is. It is the fact that Bitcoin is also an agreement. It’s not an agreement like the terms and conditions (Ts & Cs) of the software and websites we all use daily — Ts & Cs that are unilaterally and frequently changed by the company publishing the software or operating the website. There are, notably, no published Ts & Cs for running or using Bitcoin.

Bitcoin is an agreement to run software that forces all its participants to adhere to rules that arise from running this software. At a high level, the rules center around the creation of a single, indisputable chronicle of the history of Bitcoin transactions from its genesis up to the present — a chronicle that can be added to but never modified or erased. When conceptualized, these rules lead to Bitcoin’s issuance schedule and supply cap of 21 million Bitcoins and to the security characteristics that prevent it from being stolen or fraudulently spent.

Any changes anyone makes to Bitcoin software that violate any of these rules or any transactions they try to send that violate any of these rules will be rejected by everyone running the Bitcoin software. This is why it’s so hard to change Bitcoin. Unless everyone upgrades, none of these fundamental rules can be changed.

Part one of this series explained how some changes, called soft-forks — changes that don’t violate the prior rules — can introduce additional restrictions that allow for more functionality. But even those are very hard to put in place because they require all the miners of Bitcoin to enforce them and to simultaneously begin their enforcement. Part one also provided an overview of the history of how hard those changes were to make in the past.

Which leads us to the present. Does Bitcoin need further changes? What might those changes be? Will those changes take place?

At almost any point in time that one looks at there is disagreement about what Bitcoin is, what it is meant to be, and what it is not meant to be. However, at some times the discussion about this escalates towards what looks like a coming change — or at least a proposal for such change. We are at what appears to be the start of such a time right now.

Like almost every other change in the past, today’s discussions revolve around how to scale up Bitcoin. “Scaling up” often refers to how many people can use Bitcoin, how inexpensive it can be, and how fast it can be. The desire to scale up from what Bitcoin is now capable of comes from a shared vision many people have that Bitcoin should be accessible for everyone in the world, to all use at the same time, for everything. That vision requires that these additional goals be accomplished without sacrificing Bitcoin’s most fundamental tenets, like decentralization, unconfiscatability, permissionlessness, trustlessness, incorruptibility, imalleability, and so on. (Note that these big words are what sets Bitcoin apart from the existing financial system and, indeed, most other power structures in the world. Most of these words are opposites of words we use to describe traditional systems, and they have really only come into use thanks to Bitcoin. They’re so unique to Bitcoin that you won’t even find many of them in the dictionary!)

As it stands now, though, with its current rules, Bitcoin cannot achieve that vision. With eight billion people alive and transacting, there is no way under these current rules to make room for everyone to use Bitcoin. Even with just its current user base, we are seeing pressure on the cost and speed of using Bitcoin. This is no reason to lose hope, though. After all, Bitcoin is hope.

Indeed, the pressure we are seeing may be thought of as Bitcoin’s way of saying to us that it is time to figure out how to scale it to the next level.

We can’t just make Bitcoin bigger. Brute force changes, like making the block size bigger, won’t work for accommodating eight billion people — not even if we only used the block space to create lightning network channels that could each be used to transact off-chain thousands of times. And a change like this would corrupt Bitcoin’s decentralization, so it is off the table anyway for that reason. It would also require a change known as a hard fork, which part one of this series explained was impossibly difficult to implement. So it’s off the table for that reason, too.

Some new ideas even take into account the expectation that there will be many more casual users of Bitcoin who just want to use it without undertaking the effort of verifying everything. These proposals would build in protections for them even under very high ratios of casual users compared to dedicated users, making it impossible for the dedicated users to steal from or deceive the casual users.

This idea of packing lots of data into just a little space is the principal one behind all the discussions in the present. It is often referred to as increasing “economic density.” Building the applications that increase economic density is referred to as building layers atop Bitcoin. In this approach, Bitcoin itself doesn’t do all the work. But Bitcoin is an essential and necessary “base layer” that these applications rely on to operate. Bitcoin becomes the “verification engine” and even the “court of appeal” to ensure that every participant plays by the rules.

But, all of these proposals also require that Bitcoin do something it can’t yet do. It requires that we are able to restrict what can be done with the money in that transaction in one single transaction. And as part one of this series explained, adding restrictions is exactly what soft forks do. (In that article, a soft fork was compared to adding a new restriction saying “keep to the right” on a path in a lawn that already had the rule “stay off the grass.”)

Why do we need new restrictions? The basic idea is that if we want to increase the economic density of on-chain transactions then many people are going to have to share one single transaction. And if that’s the case, we need to put restrictions on such transactions so that the people with the keys to those transactions can’t steal the money that doesn’t belong to them.

Imagine if, for example, a company is making payroll to a thousand employees. Instead of making one thousand different transactions to one thousand different people, they could make one transaction for the entire sum of the entire payroll but place initial restrictions on it so it could only be spent directly to the addresses of the employees in the amounts owed to each employee — one transaction instead of one thousand. Furthermore, the idea can be enhanced by adding that when employees want to spend that money, they are able to use Lightning or some other spending protocol that doesn’t require further on-chain transactions. If they each do ten further off-chain transactions with their payments we can end up with 10,000 transactions in the space of only one. And to top it all off, if they can then keep rolling over their balances at each payroll cycle into a single new replacement transaction, we can have tens of thousands of transactions in the space of only a few. That’s powerful scaling.

All these scaling ideas remain at the pre-prototype level. Moreover, they require the implementation of two things: First, they require a consensus change “soft-fork” to Bitcoin. Second, they require the development, testing, and adoption of new types of “wallets” that utilize these changes to provide the above benefits. Both of these things are hard to complete and quite time-consuming on their own.

This creates a Catch-22 or chicken-and-egg situation in which developers are reluctant to work on the wallet software without assurances that the base layer changes will take place, but the community is reluctant to go through the effort of activating a change without the demonstrated commitment of developers to create the wallets.

Moreover, to make matters harder still, there are different approaches to the soft-fork changes that are being proposed. It is hard enough to get the whole Bitcoin community to agree when there is only change to approve, but when there are multiple proposals which are similar, and the differences between them are nuanced and not easily understood even by those who are developing with them, it gets even harder, because we can’t make all the changes. And we can even make one of the changes unless we have very broad agreement.

And that is where we are today.

To those who follow this discussion, there appear to be many divided camps: there’s those people who are very wary of making any of these changes; there’s those who are supportive of one or some proposals but wary of others (and this includes divisions that have some people supportive and wary of different combinations of proposals), there’s those whose support is conditional upon seeing some software developed first that would validate the effort of going through the change. We are confronted by a more complex situation than at any time before when changes to Bitcoin were being proposed.

Bitcoin’s governance is unlike that of companies or governments. There won’t be an executive decision made to proceed or not by any president or board of directors. There won’t be an election held. Every change so far in Bitcoin’s history has followed a different course (see Part 1 of this series), and the next one looks unlikely to follow the same path as any previous change.

Bitcoin is also much bigger today than it was during any of the previous changes. Today, there is much more capital available to invest in Bitcoin. Today, it has the attention of governments all over the world. Today, it has the buy-in of much of Wall Street as a legitimate and unique asset class. Today, there are numerous public companies involved in mining Bitcoin (and miners play a crucial part in activating changes to Bitcoin — also see Part 1 of this series for an explanation there). All these new factors may figure significantly in the complexities of how this issue evolves and, ultimately, how it resolves.

One thing is certain: this year will see a lot of discussion, debate, and disagreement on the issue of if and how to next change Bitcoin. Will we find our way to a course that makes Bitcoin self-sovereign money for the entire human population of billions of people, or will it only be an enabler of this type of self-determination accessible to mere millions who recognized its power early?

If we don’t make these (or other) changes soon, will they get harder or easier in the future?

This last question will be the topic of the final installment in this series. It will explain the trends that could lead to what is referred to as ossification, a situation in which, for all practical purposes, it will be impossible to make any such changes to Bitcoin. Why this could happen and whether or not it’s good for Bitcoin will be explored there.