By 2005 or so, it will become clear that the Internet’s impact on the economy has been no greater than the fax machine’s.
Krugman was dead wrong, but rather than mocking a man who now says he does not “claim any special expertise in technology,” I think we are better served by fleshing out the extremely apt Internet:communication::Bitcoin:finance analogy his critics raise. If we can see why Krugman might have thought that the Internet would not be significant, we might come to understand why many smart people think Bitcoin isn’t compelling and be able to explain why they’re wrong.
The Internet is a telecommunication system, but it was not our first telecommunication system. Telegraphs and telephones have been around for over a century. Like these older systems, the Internet allows us to communicate, but it differs in some important ways. Perhaps the biggest difference in the Internet model is the abstraction of a separate “application layer.” Core Internet protocols, such as TCP, part of the “transport layer,” shuffle packets of data around, but they don’t define how the exchange of packets is then used to create meaningful communication. Internet applications, such as email and the World Wide Web, are defined in protocols implemented on devices at the edges of the network, like servers and home computers, not in the guts of the network: routers, switches, hubs, and exchange points. The lower layers of the Internet can be completely oblivious to the specific applications that are in use; they just focus on getting packets of data to the right place.
In contrast, the traditional telephone system defines its applications more centrally. Improvements or new applications have to be supported by the network before they can be used. For example, the iPhone 5, released in 2012, supports a feature called HD Voice, essentially a set of audio compression algorithms that greatly improve call quality. However, this feature remains unsupported on major US carriers such as Verizon and AT&T in 2014. Because the “calling” application of the telephone system is defined centrally, the network needs to be upgraded before you can take advantage of the iPhone’s capability for crystal-clear calls.
The Internet model improves upon the traditional telephone model by making possible what Vint Cerf calls “permissionless innovation.” Tim Berners-Lee was able to launch the World Wide Web without waiting for Internet service providers to support it. Innovative Internet-based voice and video communication services like Skype, FaceTime, and Hangouts work as long as users on both ends use the same software. If Sir Tim had to explain to a telecom executive what hypertext was in 1990 before he could create the web, it may never have happened. If we had to rely on telecom companies to provide video calling (AT&T experimented with it as far back as the 1960s), it would probably be more expensive and inferior to the video calling services we have today. Permissionless innovation means more innovation.
Moving decision-making to the edges of the network also has the effect of decentralizing power. End users can decide for themselves whether to use encryption, whether to engage in the sharing of copyrighted files, and indeed, whether to use a cryptocurrency to buy drugs on a Tor hidden service. Notwithstanding the Snowden revelations, Chinese censorship, and the occasional use of Internet kill switches by regimes facing domestic uprisings, the Internet represents a great devolution of social power. Centralized networks have never supported the ability to easily break copyright and drug laws or to engage in technical measures to prevent government eavesdropping; they keep social power in the hands of the businesses that run—and the governments that regulate—the network.
It is easy to see how Paul Krugman, someone without any special expertise in technology, writing in 1998, could miss the importance of the Internet. The permissionless innovation and devolution of power fostered by the Internet may be obvious, or nearly so, in retrospect, but not in advance. From Krugman’s perspective, perhaps, the Internet was just another medium for communication, of which we already had plenty. Why should it have been such a big deal?Which brings us to Bitcoin. It is a currency, of sorts. You can spend it on things, especially drugs and gambling and getting around capital controls. Krugman and other economists have analyzed Bitcoin in these terms, as a substitute for dollars. This is rather like regarding the Internet as a substitute for, and not a quantum leap beyond, previous communication technologies. It is true that Bitcoin can substitute for other currencies, but as with the Internet, the abstraction of a permissionless application layer means that it is much more than a substitute: it is like a transport layer for finance.
Every Bitcoin transaction is defined in part by a bit of code, called a script, written in a programming language called Script. The script in one transaction defines how the next user can access the coins. In a conventional transaction, the script specifies the hash of the public key that is needed to spend the coins next, and demands a signature from the corresponding private key.
Script is not limited, however, to these conventional transactions that merely transfer coins from one person’s control to another’s. It can evaluate statements, execute conditionally, do math, and move bits around. It is not a Turing-complete programming language (there is no looping), because that would be a security risk; we do not want viruses to spread via Bitcoin’s blockchain, nor do we want Bitcoin transactions to run indefinitely or, if we ever figure out AI, become self-aware. Despite the lack of loops in Script, it can be used to construct some very interesting scripts. As I have previously blogged, multisignature Bitcoin transactions can include a built-in arbitration component in case of a dispute. This is a big deal, because competitive arbitration could give rise to an emergent law of Internet payments that is completely independent of state-run courts.
Script can support even more complicated, multi-party transactions. For example, assurance contracts, in which people agree to fund a project conditional upon others contributing a specified sum, can be coded in Bitcoin. Alex Tabarrok’s variant, dominant assurance contracts, which pay back an extra penalty to contributors if the project doesn’t pass its threshold, can also be scripted. Bitcoin scripts can also consult an oracle, a trusted computer program that publishes information into the blockchain, Bitcoin’s distributed ledger. This enables information about the external world to be embeddable into Bitcoin transactions. The decentralized prediction markets that Ed Felten’s team at Princeton is working on are based on this principle. In short, Bitcoin is not just a substitute for money; it can be a form of generalized, programmable, decentralized contracting. Legalese gives way to Script.
Another feature of Bitcoin that hasn’t been widely used yet is the micropayments channel. This allows what are essentially continuous payments, making fine-grained usage-based pricing possible for essentially the first time. Some members of the Bitcoin community think this will finally make wireless mesh networking a feasible proposition, since it will eliminate the incentive to free ride by refusing to relay data. Continuous payments would be both sent and received automatically by your phone or laptop’s OS, netting out, providing cheap (or, on net, even free) universal, seamless wireless Internet access. Since continuous micropayments have never even been possible before, it remains to be seen what other applications they might have.
Bitcoin also has applications that are not purely payments. As a distributed ledger, it can function as a notary service, providing proof that a given document existed at a particular time and date. Another innovative idea is to use Bitcoin as a bonded identity service. The proposal is to create a system of “Passports,” secure identities verified by the blockchain, backed by “fidelity bonds.” Essentially, you would create a unique identity in the system and then verifiably give a non-trivial sum of bitcoins away to whichever miner randomly processes your transaction. By making it costly to create an identity, this system could be used to ensure that dozens of new accounts are not simply created to engage in fraud, spam, or sock-puppetry. Passports could be portable across different online services, creating a persistent identity that is both pseudonymous and reputable. Reputation rating services could develop to ensure that Passport-holders who misbehave on one service are shunned on other services.
In contrast to these possibilities, the existing financial system is terribly boring. You can use systems like SWIFT, ACH, or FedWire to move money around, but you can’t exactly use these to do anything else. The comparison between these legacy systems and the old, pre-Internet telecom environment is apt. To do Bitcoin-like innovation without cryptocurrency, you need to convince a whole network of banks to adopt your system. It is not permissionless; it is prohibitively permissioned. Bitcoin is a new transport layer for finance that allows decentralized, disruptive, permissionless development of applications on a separate layer. It has the capability to do for finance what the Internet did for communication.
In the course of developing these new applications, social power is likewise going to be redistributed. Just as the old telecom system concentrated power in the hands of the telcos and the governments who regulated them, the existing finance system concentrates power in the hands of banks and governments. If Bitcoin or another similar cryptocurrency succeeds, they can kiss that power goodbye. Not only will it no longer be possible to censor simple transactions, such as donations to WikiLeaks, it won’t be possible to ban or regulate certain classes of transactions and contracts, such as betting markets and assurance contracts. The power to decide whether to engage in these activities will be redistributed to ordinary people.
Most of Bitcoin’s critics are making a category error. They are taking aim at Bitcoin the currency, when in fact Bitcoin is much more than a currency, in the same way that the Internet is much more than the telecommunication services that preceded it. It may take some time, but we will likely one day look back at the critics’ current statements with as much derision as we do a comparison of the Internet to a fax machine.