One Cryptocoin To Rule Them All
But the above still does not address the main point that Krawisz makes – namely, that because cryptocurrencies benefit from network effects, there must necessarily eventually be one cryptocurrency dominating the market. In general, the principle that network effects create a tendency toward monopoly is a valid one. The core idea is that some goods (or standards, platforms or protocols) are inherently more useful if more people use them; for example, a radio would be essentially useless two thousand years ago simply because there were no other people with radios to talk to. When applied to protocols and standards, this means that a standard that gathers more users immediately becomes more useful by virtue of having more users alone, and so even more users adopt it, creating a positive feedback loop that ultimately culminates in one or a few standards establishing market dominance.
We can see this effect in the real world in many places. There are still hundreds of languages in widespread use today, but the reason why this status quo survives is that there is still not yet much of a global network for network effects to work off of; by and large, Chinese people talk to other Chinese people and Americans talk to other Americans, so different languages can come to dominate in different regions. In areas where this is false, like aviation and international business, English dominates worldwide. As another example, Krawisz writes, “Suppose automobiles had just been invented and two groups, because of vested financial interests, got into arguments about whether it was better to drive on the left or the right side of the road.” For obvious reasons, every such conflict in the real world has now been resolved; left-side driving only remains in highly contiguous pockets of the world – that is to say, self-contained regions where network effects (ie. people driving across borders) are small enough for the discrepancy to be tolerable. In the area of social networking, Facebook remains by far the most popular of its kind, although its importance diminishes somewhat if you include into the category of “social networking” things like Twitter, online forums and email.
In the context of alternative cryptocurrencies, however, the last example is particularly pertinent. It is virtually guaranteed that no company will be able to “beat Facebook” by simply being a different version of Facebook – as opposed to restaurants, where network effects are virtually nonexistent and anyone can set a new restaurant up and be moderately successful without offering any particular innovation. However, the services that have survived Facebook are those that can work alongside it, and have different properties that make them far superior for some particular use cases; Twitter and email are the obvious examples.
Perhaps, it will be the same with Bitcoin. Bitcoin itself is not nearly perfect for every use case; transactions take up to sixty minutes to confirm depending on the desired level of security, the system is not completely micropayment-friendly, and its scripting system is very limited compared to what it could be. Imagine, for example, a Bitcoin address that allows one private key to spend up to 0.1 BTC per day and the other to spend everything. Such a system would be a boon for security and applications like recurring billing; however, the current scripting system does not allow reading transaction values or past transactions. A future coin might. Ripple allows users to set up their own currencies and includes a decentralized exchange; Bitcoin does not. Bitcoin’s slow confirmations are already solved by Primecoin, whose blocks take one minute to confirm. Bitcoin’s lack of ability to (easily) include messages along with transactions is solved by Neocoin.
But then, the obvious question is, why use altcoins for all these things? Why not integrate advanced scripting, rapid block confirmations, lower transaction fees and messages into Bitcoin itself? In theory, it could be done. In practice, however, there are several reasons why it should be done and several reasons why it simply will not be done. First of all, complexity is bad. If nothing else, the blockchain fork in March 2013, where the Bitcoin network literally split in half for six hours because of a discrepancy between two pieces of database software, made this all too clear. If the Bitcoin network can be brought to its knees by an obscure database limit that Bitcoin developers did not even know existed, then imagine how well a version of Bitcoin with dozens of interlinking parts and dependencies would fare. If advanced features are implemented in an altcoin, however, the risk of failure is lower in each one, and if an altcoin falls down due to a technical glitch there will be far fewer people using that coin to store millions of dollars or make business-critical financial transactions who stand to suffer. Even if one does not take into account the risk of failure, complexity is harmful simply because it makes it more difficult to create alternative implementations of the software, stifling innovation in software development and leading to a kind of centralization by the backdoor.
Second, there is the matter of resources. Consider the case of Neocoin. Many Bitcoin developers are dead-set against integrating messages into the blockchain; when blockchain.info tried to implement such a system using zero-valued outputs, a number of Bitcoin developers, including Gavin Andresen himself, heavily criticized the initiative. Bitcoin developer Luke Dashjr wrote: “Data does not belong in the blockchain. People running nodes have all implicitly agreed to store the blocks for financial purposes, and storing data is a violation of that social contract.” The fear is that if the Bitcoin blockchain becomes recognized as a place to store messages, and people start using it as an email and instant messaging (or worse, file sharing) service, then network load would balloon massively and the blockchain would grow to many terabytes in size. With Neocoin, if that happens no one would have any cause to complain; the community would work on reducing the storage requirements, including potentially modifying the mining algorithm to remove the need for every miner to maintain a full blockchain. Similarly, Zerocoin, with its multi-kilobyte anonymous transactions, can grow to function as an effective mixing service on its own currency without saddling the Bitcoin network with additional data or the Bitcoin source code with advanced mathematics that even most professional cryptographers do not understand. Sure, the total amount of data would be the same, or even greater due to the need to exchange between the different currencies, but each individual would only need to worry about those blockchains that are relevant to them.
Switching Costs Go to Zero
All of the above arguments show the advantages of having different blockchains. But there is also a strong argument to make that the disadvantage of having multiple currencies is really not all that great. Krawisz cites several examples to show how competition between standards can be bad; aside from the example of driving on the left and right side of the road, he asks: “do we really need competition between the mile and the kilometer, for example?” The fallacy here is this. Miles, kilometers, traditional currencies and road regulations are all standards meant to be used by humans. Faced with a dual-measurement-unit system, we as individuals need to keep track in our heads the fact that one mile is 1.6 kilometers and one kilogram is 2.2 pounds, and use mental arithmetic to convert between the two as needed. In the case of the dual standard of “miles per gallon” and “liters per 100 kilometers”, the situation is even worse: one is an inverse of the other, so MPG going up is good whereas L/100KM going up is bad. If different cities had different road regulations, not only would drivers need to remember which side of the road to drive on in each city, but every boundary would need to have some kind of special streetlight or crossover bridge to allow drivers to switch from one to the other in real time.
Cryptocurrencies, on the other hand, are different. Because the actual handling of currency is done in software, bridge applications that allow people to send one currency to an address in another currency through an exchange can easily be created if necessary; Ripple recently introduced a tool that allows Ripple users to send directly to Bitcoin addresses by using the gateways in the Ripple network. Multi-currency wallets are also another possibility. Altogether, we have the tools to completely abstract away currency divisions from the user, and treat them purely as a technical problem; users need only choose the unit of account to store their money in, and peer-to-peer decentralized exchanges can take care of the rest.
Bitcoin itself suffers from a network effects deficiency against fiat currencies; currently, everyone uses fiat currencies, putting Bitcoin at a considerable disadvantage at least in these early stages of its growth. Bitcoin has only been able to overcome this problem simply because there are some circumstances in which it is simply so vastly superior that even the massive network effects of the euro and US dollar were not enough to overcome its advantages. Once Bitcoin wins, Krawisz argues, altcoins will be in a similar situation. “To defeat Bitcoin, an altcoin would require not just superior technology, but such vastly superior technology as to be an advance over Bitcoin comparable to the advance Bitcoin represents over fiat currency,” Krawisz writes. However, this argument is false. The reason is that network effects are not the only factor driving standards toward monopoly; the other major factor is switching costs. Making the switch from fiat currency to Bitcoin is quite difficult, and creating bridge services that work between the two is a multi-million dollar affair involving placating established banking partners and complying with a large body of regulation.
The path from one cryptocurrency to another is much simpler. Anyone can set up an anonymous exchange over Tor, and it may even be possible to make trust-free decentralized exchanges between currencies using Bitcoin’s underlying cryptography directly. For a merchant accepting Bitcoin, if alternative currencies gather any popularity at all then switching to Litecoin or Primecoin will be a simple matter of downloading a patch. It is even possible for a merchant to accept every cryptocurrency at the same time. Certainly, a new currency will need to be substantially better than Bitcoin to overtake it, and Bitcoin’s status as the first cryptocurrency out there gives it special unique value (much like “the original” of any painting is worth thousands of times more than even the most visually indistinguishable replica). However, as far as adoption is concerned, it is perhaps the chief benefit of cryptocurrency that it massively reduces the height of the “hump” that every new currency innovator will need to overcome in the future.
