Why Global Trade Will Inevitably Move to the Blockchain
As part of Blockstack’s “Decentralizing the World” Tour, I was invited to speak about blockchain and global trade. You can see the video above, and below is a transcript of the talk.
The situation that we’re in right now is almost all of the explanations for why the blockchain is interesting or necessary are total hogwash, they’re just nonsense. Because they refer to a bunch of cultural factors or technical factors, but they don’t refer to the underlying scientific and mathematical factors which all of this bedrocks onto.
Has everybody heard of a thing called the speed of light? Raise your hand if you have any practical, working relationship with the speed of light on a day-to-day basis. You are all wrong! Have you ever used GPS? It’s a bunch of clocks floating in orbit around the Earth, and the speed-of-light delay between you and that clock differs depending on where you are relative to those satellites, and that delay is what creates the triangulation that makes GPS work. GPS is based on speed-of-light signal delay, and it’s enough signal delay that $100 phone is capable of positioning you within a couple of metres. So the speed of light is so slow relative to modern computing devices that it creates an easy way of manipulating time and space in completely revolutionary ways.
However, it also means that it’s impossible to make all the world’s computers agree about what is real, because they’re all separated from each other by the speed-of-light delay, and the speed-of-light delay say between London and Australia is roughly a million transactions delay relative to modern transaction processing systems, so the machines around the world are separated from each other by a million transaction delay, and that fact of life is never going away. Physics does not permit any closer synchronisation than we currently have, you could tweak the engineering a little bit with fibre optic cables or microwave lengths or however you like, but we have no way of synchronising the computers. So if you want to use computers to figure out who owns what, unless you have some way of working around the speed-of-light delay, you’re going to wind up with no synchronisation, which means you’re going to have continuous fighting about property rights. “Well, I say it belongs to me, but he says that I sold it to him, but I actually sold it to this guy over here.” Because I’m in London, he’s in France and he’s in Australia, and I can just continue to spend transactions, millions of them, while I waiting for the signals to propagate.
For this reason you need something that figures out how to manage the speed-of-light delay. Now, the model that we’re currently using for this is called high-frequency trading, where you have a single exchange that moves the asset that you’re trading, and that single exchange measures out the cables with a ruler to the closest millimetre so that everybody has exactly the same speed-of-light delay, and then the techs begin to argue about whether the cables are equally straight. The bottom line is that that might have worked when you had undisturbed American hegemony, where they just ran the world their own way and the exchanges were in America, but in a multipolar world why would the Chinese trade all of their commodities on American exchanges? And if they trade them from China and they’ve got a China-to-America speed-of-light delay, how are they going to operate a million transactions behind the American systems? You see how there’s just no way this works? Yeah, that’s because there’s no way this works.
In all probability, the geopolitical equilibrium that we go into has a large exchange for China, a large exchange for America, a large exchange for Europe and maybe a few others. It could also turn into a couple of thousand exchanges which do individual commodities or they’re city by city. But one way or another, one exchange for trading 50% of the world’s commodities is not the way the future looks, for political reasons. And then you have a choice, which is do you decentralise all the way, or do you partially decentralise and then use some way of synchronising the exchanges? The only way that we know of synchronising this stuff in any kind of sensible, reliable way is to basically stop time. You take all of the transactions that are plausibly happening inside of a given window, you then stop the clock, clear all of those transactions and make sure they don’t contradict with each other, and then you start the clock and you do the next batch of transactions. We call those batches blocks. Speed-of-light delay is maybe a quarter of a second up to half a second, depending on the Internet’s mood that day, Bitcoin block time is 10 minutes, Ethereum is 15 seconds. That 15-second batch gives you the ability to clear all of the transactions and make sure that everything is exactly where you think it ought to be, before you go on and do the next round of trade.
We’re not going to do better than that with current physics. It’s just possible there’ll be some kind of algorithmic breakthrough, but in all probability we’re not going to do better than that with current physics. So there’s going to be something that looks kind of sort of like the blockchain; at the very least you’re going to have transactions, stop the clock, clear the transactions and start the clock again type systems. Maybe there are some refinements, maybe you could get the delays down to a couple of seconds, but, in all probability, this is the inevitable equilibrium, because there is no other way to synchronise the world’s computers.
Let’s then talk about how this impacts global trade. As we stand right now, you’ve got on the order of $100 trillion a year of GDP globally, and about half of that is physical goods in the industrial supply chain. For size of scale, ecommerce is about a trillion dollars a year, so we’re dealing with an industrial supply chain which is 50 times the size of ecommerce. About half of those commodities are going over what they call high-frequency trading, which is the big exchanges in America, $25 trillion a year ballpark. Now, in that sort of setup, what percentage of the friction involved in doing international trade is just nonsense, crappy paperwork? Has anybody dealt with international payments recently that weren’t made using Bitcoin? Has anybody tried to ship anything remotely unusual across the world? Has anybody tried to buy electronics from China? It all sucks! Things get stuck in customs, the same thing ordered from three different days gets stuck in three different ways, you have variable fees that get charged at random and your stuff gets stuck for weeks. If you’re dealing with container shipping it’s even worse: your box is stuck in Singapore, nobody seems to know where it is, eventually somebody discovers that that the paperwork was filled out in Cantonese rather than in English, and as a result the thing has stalled, etc. The physical world is this nightmare of randomness and incompetence, and all of the randomness and incompetence in the physical world wads up against the paperwork and then sticks there.
When we came to container shipping, where we go the monkeys out of the loop and we’re no longer carrying huge bags, we’re moving things off and on trucks, when we automated all of that stuff with what looked like packet-switched networks, it put, best estimate, 7% on global GDP, 7%, $7 trillion a year by modern numbers. So what can you do if you managed to kill all of the crap around invoicing, international payments and supply chain management of goods? It’s at least 2–3% of global GDP, it could easily be 5% or even 10% of global GDP. And that’s before you begin to think about the efficiencies of getting rid of factors like trade gravity, so that people start to trade with who it’s more efficient to trade with rather than who it’s easiest to trade with. The more you bring down the non-tariff barriers, the more you solve what they call the trilemma of globalisation, the more you handle those problems, the more it liquidises global trade so you trade with the most efficient partners rather than the most convenient partners.
All of that is a technology problem. We’re not going to leave global trade unautomated, in a pre-Internet condition. Inevitably, a hundred years from now, there will not be paper bills of lading, 20 years from now there will probably not be paper bills of lading. The question is six months or a year or two years or five years from now will there still be paper bills of lading? You have a curve: if things go quickly, we get all of that stuff very quickly; if things go slowly, it takes much longer to get rid of it. But somewhere in the next decade to two decades, all of that stuff is going to flip digital, and it’s going to inevitably wind up on something which is basically blockchain flavoured, and the result is going to be a multipoint bump in global GDP.
This is going to step directly into the gap which is being left by the breakdown of the international political order. You take institutions like the European Union or NAFTA or the Trans-Pacific Partnership or the World Trade Organisation, all of these things are coming under increasing amounts of pressure from populations that have basically just had enough of industrial-age globalisation diluting their democracies, and gigantic corporations which have had enough of governments trying to tell them what to do in a one-size-fits-all way. Everybody wants more control of their lives, and the only way that you could get efficient control of an enormous amount of diversity is with flexible and intelligent computer systems. We’ve got no problem providing every individual on Earth with whatever it is they want to read on any given day for a very, very reasonable price, because we figured out how to provide completely flexible, mass-customised access to information; you just figure out what you want to look at on the Web and it’s there.
We could do the same thing for material goods, if we could just get the shipping right to move the things around, and the bureaucracy right so that when you decide you want to buy something from New Zealand and you happen to be in Belize, systems that are as good as the Internet is at routing information route to matter. You just connect the pieces together: that tariff is over here and this customs agent is over there, and this goes over here and this goes over here, a bunch of computers talk to each other, and then a bunch of magic happens and tells you it’s going to cost you a $1.80 for charges and ship the thing now. We could get to that world from here, and blockchain is the ideal way of coordinating all of these decentralised pieces of the global bureaucracy together into something that performs as well as packet-switched network.
Now, getting that to work is going to be, on one hand, a sodding nightmare, it’s going to be really difficult, because it’s bureaucracy. On the other hand, once you begin to get the knack of automating bureaucracy, and using blockchains to automate bureaucracy, and you go from smart contract to smart statute, and you begin to represent things like customs regulations at ports as a set of interlocking smart contracts rather than as a bunch of paper documents, once all of that stuff comes together, what you get is potentially the single largest jump in global economic efficiency since industrialisation. The Internet has not even remotely began to impact the global economy yet. We are still at the point where we’re basically just organising the libraries, and figuring out what to do instead of catalogue shopping and magazine shopping. If I want to get 100 kilowatt diesel generator delivered to Alaska tomorrow morning, I’m still going to wind up on a telephone. So we haven’t really began to touch the real economy yet, and this is also part of why we’re able to play fast and loose with the law.
As we begin to get into the serious challenges of doing this kind of enormous heavy lifting which is required to move the world’s assets around, so that you change the status of objects using a smart contract and the legal ownership immediately changes, the contractual status of an object changes, once you get to a point where the point-and-click interface to the legal reference layer that defines ownership and property, the size of the jump in economic efficiency could be genuinely titanic. I think probably 20–30% of the assets of the world are shipping around underutilised, because we can’t efficiently get the things to market. You just look at what’s happened with Airbnb and with Uber to get a sense of the scale of these transformations
Within that big story there are some simple starting points. There must be 50 companies working on blockchain bills of lading, there must be any numbers of people working on port automation, there are lots of people working on things like digital identity. Mattereum, my company — my co-founder Rob Knight is over there — we are sitting there, doing the dog work of trying to figure out exactly how to get control of physical assets and intellectual property from a blockchain, so when you transfer it a smart contract the court agrees it’s transferred… There are lots of people chipping away at different parts of that problem, but the long-term impact of that is basically direct digital control of matter as a daily fact of life, and that’s what we’re building. Information was only the first thing to get digitised; now we’re going to do the digitisation of matter. Matter will become searchable, matter will become indexable, matter will become something that you provision exactly as we need it, and it will be delivered as quickly as information or damn near it. And that doesn’t have to be a kind of one-size-fits-all, Amazon rules the world approach, because there’s too much matter and too diverse configurations for a single bureaucratic approach to tie the entire thing together.
What you need are these kind of decentralised backbones, you need protocols, but you also need decentralisation of the software that implements the protocols, and what comes out of that is this revolutionary, almost imaginable change, which is 20 years from now matter will feel like information does today. Because information feels like it does today in this cheap, super abundant, liquidly-accessible form, because we managed to get it under the proper digital control. If you go back to the 1960s, 1970s, 1980s, pre-Internet, information was incredibly difficult to access, it was just hard. It was expensive, it was difficult, there were huge overheads and it was hard to move it around. If you’re young, you think that information was always this kind of hyper liquid commodity that flowed around the world like air. That’s not an inherent property of information. The machinery we built to revolutionise the world worked first on information, it worked best on information, but with the right kind of diligent effort, and blockchain is a critical part of that, we can drag the matter along with us into a similar state of hyper liquid accessibility. I can give you plenty of examples of how that might work, but the basic story is first we did information, next we’re going to do matter. That’s what I had to say — thank you! [applause]
