Single Global Currency
In 2009, the year of the Bitcoin paper, I made a short slide deck and met with some board members and execs about how the Second Life Linden Dollar, which was rapidly approaching a $1B economy, might be made into a decentralized global currency.
10 years later, cryptocurrency still hasn’t yet become useful for ordinary transactions, the LindenDollar economy is similar in size but hasn’t grown, and Facebook has just proposed Libra. Some thoughts and comparisons, looking at this deck and then back at these first 10 post-Bitcoin years:
Better than Cash
Inside Second Life, two people from anywhere in the world (meeting as avatars) can instantly pay each other for anything. You can give a tip to a musical performer or buy a t-shirt for your avatar. It doesn’t matter what countries the two of you are from, and the fees are either zero or very low. The whole point of Linden Dollars in Second Life was to make something as easy to use as cash in the real world, but for avatars. And so naturally, first and foremost, I was interested in the question of whether some sort of digital currency on a smartphone could replace cash for actual people from different world regions. Now to replace cash for a consumer, you minimally need three things: a fairly stable currency, low or zero transaction fees, and a way to easily transfer money to a newly-met person or business. In the last 10 years we’ve seen a lot of interesting progress with cryptocurrencies, but none that have yet met all three of these basic requirements. In the specific language of crypto, what we need is a stablecoin with low fees, combined with a universal wallet that makes it easy to send or receive money. And no one has done that with any real success, yet. The closest thing I’ve seen recently is the Coinbase wallet using the USDC coin. But that is still a custodial, proprietary wallet, and the fees are only low if you both have Coinbase accounts. In the non-crypto world you have things like Venmo in the US or WeChat in China, but these are restricted to citizens of those respective countries. So we’re not there yet.
Legal & Regulatory Hurdles
Understandably, if a company tries to make something like cash, there are big problems to contend with, for several different reasons: Countries don’t want to lose out on tax revenues, or the ability to manage their own currency. Police forces are rightly concerned about use for money laundering, funding terrorism, or buying illegal goods. Banks and credit cards don’t want competition, and will therefore lobby to erect the highest regulatory barriers possible to new entrants, as well as denying access to their own services.
Furthermore every country and/or state has different rules and opinions about all these things, which was something we got to see close up when creating Second Life. Because avatars came from all over the world and were constantly interacting both socially and financially, we got a painful lesson in how impossible it was to come up with policies that made every country happy. Based on this experience, I can say with some confidence that it is going to be nearly impossible to do something in the way of a new currency if you a-priori need multiple countries to agree with and support the idea.
So, for that reason, I started in 2009 trying to think of how to make some sort of ownerless system where lots of different countries and institutions could easily opt into initially being part of it, with the hope that once the system got started the cost of any one party opting out would be less than the benefits of staying in. My plan (from the slides) was to create a network of ‘balance servers’ that would stay in sync with each other using a basic consensus algorithm — the same sort of strategy Facebook is proposing for Libra.
“Money is never created or destroyed”
This cryptic sentence on slide 3, combined with the suggestion of open exchange to fiat, was my hand-waving non-answer to the question of how to generate units of the actual currency, and I can only defend myself by observing that in the past 10 years nobody else has made much progress either:
The most pragmatic solution, which is what Libra is proposing, is to trust the people running the servers to accept deposits of real currency in exchange for creating units of the digital currency, maintaining a 100% reserve so that if everyone holding the digital currency wanted their money back they could get it immediately. This will work fine and of course be stable, but you have to trust the people running the servers to mint new currency only for real deposits. Since a currency like Libra is a public blockchain and reserve funds easy to audit, this really should work just fine for the basic goal of enabling payments with a stable value. There are some complexities around how to make a ‘basket of currencies’ work without enabling various types of triangular arbitrage, but it can probably be made to work.
A different solution is to create units of currency that are given out in exchange for some provable event or action, without any reserve or promise of future value. Bitcoin mining is this: the conversion of electricity (which is wasted as heat and therefore unrecoverable) into bitcoins. One would have hoped (perhaps Satoshi did) that this sort of thing might create a more equal economic system, with small amounts of currency being doled out a lot of people in equal amounts who can then go forth and use it for transactions. Unfortunately, the fixed rate at which bitcoins are mined combined with rampant speculation as to their long-term value ended up doing the opposite and concentrating coins among a small number of accounts. Another very serious problem is that heating the earth’s atmosphere as a means of generating currency is absolutely unacceptable. Even Black Mirror has nothing on the dark irony of this being perhaps our final ‘Bonfire of the Vanities’.
Second Life, by comparison, does something similar to what real-world countries do, which is to gradually create new units of currency and then either sell them to anyone willing to buy them on the open market, or give them away to stimulate growth (similar to government work programs, or universal basic income). The company does not promise to buy back the currency, so again there is zero reserve. But because there are lots of people making and selling interesting things in the world without practical alternatives for payment systems, there is plenty of demand for the currency, keeping its value steady. Obviously if people in Second Life stopped making valuable things, the value of the currency would fall. But this is also true of countries — ultimately you have to create value to keep your currency stable.
Which of these approaches to managing a digital currency is best still seems difficult to judge, in the same way there is perpetual debate over real-world monetary policy and whether things like ‘quantitative easing’ make sense. One thing which seems clear is that the Utopian desire to somehow use a newly-minted currency as a great equalizer or Robin Hood seems unlikely to be easily achieved.
15 Seconds or Bust
I think I was right to focus on easy transactions, as it is still one of the blocking problems for crypto. Quick! Give me 0.001BTC in the next 15 seconds and I’ll give you a cold beer… not doable yet. My idea to fix this was to use a 4-digit alphanumeric short code as a short-term proxy for a transaction. You enter a specific amount you want to give or get and the app gives you back a shortcode, like “X75F” that you can then show or verbally tell the other person. By making the codes expire within a couple minutes, it turns out you can cover as many transactions with 4 digits as humans can possibly do per day. It might be faster to use QR codes, given that cameras are now ubiquitous, if you can get your two phones close together, but the short code idea would also work for web payments (you’d get one from your phone and then enter it on the website for payment — with the added benefit of two factor security). But whatever strategy ends up getting used, it has to be as easy to pay someone as getting cash out of your pocket. And it has to work for everyone, not just if you both have iPhones. Now the crypto designers out there will quick point out that for this to work, you would minimally have to trust the short code generator to not cheat you, and this nicely moves us to another important observation:
It would be easy, if you could just trust someone
Bitcoin and Ethereum have confused us all by making us think that is is actually a hard problem these days to record thousands of payments per second (that’s what VISA handles). But ask any back-end server developer what kind of tech and hardware would be needed to record a thousand transactions per second into a database, and they will tell you a single Amazon AWS machine running an off-the-shelf database can easily do the job, and that the bandwidth and server costs will be at most a couple thousand dollars per month. So what’s the big problem? Distrust is the problem! If you tell that same back end developer that instead you want to build an open financial system where no one trusts each other and all the servers are presumed to be actively trying to steal from each other, the final equipment and bandwidth costs skyrockets until you have something like Bitcoin — a network consuming as much energy as a country and capable of only a few transactions per second. In theory it is possible to build a trustless, permissionless currency, but in practice it is incredibly expensive. Whether the costs of a permissionless system can be gotten down low enough remains a problem for the crypto community to continue trying to solve, but I wouldn’t hold my breath. And that is why Facebook proposes a permissioned network for Libra to start — not because they have some bad intent — but because at this point there simply isn’t a design that could serve as an alternative and handle 1000+TPS.
For me, all I want is a global-scale virtual world where people can do anything, including buy stuff from each other, safely and securely. That’s what I have been working on my whole life. And there are plenty of really hard problems with building that place, independent of payment systems. But this currency thing is one of them, and for the benefit of us all we need to get it solved. VR is a use case where the benefit to society of such a system is very positive, so long as we focus on applications for it that bring people together in virtual worlds: As we saw first with Second Life, people can get new jobs as avatars, make new friends or even life partners across borders and cultures, and stop using carbon-wasting airplanes to travel for business meetings or conferences. These are very positive uses of technology which require a working payment system to fully activate.
In preparing this blog, I’ am thankful for contributions from my co-founder and economist Freidrica Heiberger, as well as former Linden Ginsu Yoon, who oversaw many of the monetary policy decisions we made at Second Life.