It has been exactly three months since I started buying bitcoin. From my first purchase, I bought 0.06 coin at 1,160 yuan per coin; at the lowest point I bought 0.5 coin at 372 yuan. All told, my cost basis is roughly 650 yuan. Then I spent 5 coins buying mining machines, made 5 coins from stock trading, and between assorted low buys, high sells, and gambling, my overall balance of income and expenditure is about even.
During this period I wrote a series of bitcoin-related articles, and also started a column on Babbitt, which has circulated fairly well in the Chinese bitcoin community. At the same time, I also began following a great deal of bitcoin-related information on Weibo, and gained quite a few followers.
But among my old friends and old readers, there still seem to be very few whom I’ve successfully talked into joining the bitcoin world, and that makes me a bit disappointed…… In any case, buying bitcoin is on the one hand a hedging investment, and on the other hand an academic practice. This practice can connect up the things I’ve learned from history of science, philosophy of technology, media phenomenology, and related fields, which makes me personally very excited. Of course, if you lose money, don’t blame me. I’ve always only suggested taking out a small, inconsequential slice of your spare cash—say 1% to 10% of your savings—so that even if you lose it all, it won’t be a crippling blow.
Right now bitcoin’s market price is below 100 dollars (note: this article was written in fits and starts over the course of about a week, and it is still that price…), which means it is quite a bit lower than my cost basis of 650 yuan. So some people may say: good thing I didn’t listen to your hype before, otherwise wouldn’t I have taken a loss? But that’s not how things are. On the one hand, only by starting to hold bitcoin can I carry out a series of further explorations and practices. Stock, gambling, futures, and so on—if you only watch from the sidelines, it is always hard to develop a deep understanding. (I’ll introduce these one by one later.) On the other hand, my cost basis of 650 yuan was not invested all at once; it was accumulated gradually, buying a little whenever the price dipped a little. In any investment, you cannot enter with your full position at once, and even less should you hold the mentality that I must grab the very lowest price. Buying in batches, leaving room most of the time, and being mentally prepared for the risk of decline—that is the necessary attitude. If it keeps falling now, I will still happily keep buying.
For beginners who have not yet bought coins, falling bitcoin prices are actually a better thing, because it means our entry cost is lower. Keep in mind that I am, after all, advocating that everyone buy bitcoin, not that everyone make renminbi by way of bitcoin. The standard for judging whether an investment is successful is the increase in bitcoin; it is important to get this straight.
Measuring returns in bitcoin will first of all dispel the misunderstanding about “mining machines.” For many Chinese people, the first thing they do when investing in bitcoin is not hoard bitcoin, but buy mining machines, and that makes me extremely anxious. In any case, if I’ve talked you into becoming interested in bitcoin, then I must responsibly tell you: do not buy mining machines now. Most mining machines on the market are not worth buying; you don’t even need to listen to the quote to know you absolutely should not buy them, whether spot or futures—because there are too many newbies in China driving the price up, and profit-seeking merchants can’t possibly have a charitable impulse to dump them cheaply on you, unless you happen to buy, through personal connections, someone else’s machine that has already broken even and is being sold to you half-gifted out of kindness, or unless, like me, you buy a tiny USB miner for the sake of fun, practice, and showing off.
Below I’ll talk about my understanding of bitcoin mining, and along the way discuss a few altcoins (NMC, LTC, PPC, XPM), because what altcoins mainly do is also adjust the mining model.
1. Mining: maintaining the reliability of the network, and by the way, the unreliability of buying mining machines
Why are mining machines not worth buying? What is a mining machine? Let me explain a little. Of course, for a detailed introduction, please search on your own; I’ll only talk about the key points as I understand them.
First of all, the essence of bitcoin is a decentralized bookkeeping system. Every full bitcoin wallet stores all transaction records since bitcoin was invented—at present this has already reached a huge 10G, though ordinary users can choose a simplified version without full data, or even an online wallet (such as blockchain). Only I, possessing the key to the corresponding address, can use my own balance to send currency to another address. The so-called sending of currency is not sending a segment of data one-to-one, because if I only send data to the recipient, then I could also send the same segment of data to another person at the same time. Who would confirm that the first segment of data is valid and the second invalid? In other words, after I send a sum of money, the recipient should know that I have the ability to send that money, but everyone else should also immediately know that I have lost the right to send that same money to them. How is that done?
Centralized currency is simple: you only need a core authority to audit every transaction. But how do you do it in a decentralized system? Bitcoin’s solution is for every transaction to simultaneously broadcast information to the entire network: “Zhang San sent X coins to Li Si.” Then everyone else in the network must immediately do two things. First, confirm that this message really was sent by the person who has the corresponding key. Others can verify, without knowing the key, that the information I sent was indeed decrypted through a specific key; how this is achieved is a problem in cryptography, and I won’t go into it here. Second, confirm that my account really does have sufficient balance, which is inferred from all transaction records ever made. If it passes verification, then the entire 10G ledger will simply gain one more entry.
But any currency has to face one problem: preventing forgery. If I send more than my balance, and my accomplice cheats by helping me confirm that the send is valid, then what? Obviously, in a centralized system, we can hand the right of judgment to some authority, but under a decentralized network structure, no one has absolute judgment power. When some say it is valid and others say it is invalid, the final winner can only be the “majority.” If more than 50% have verified it as valid, then those minority anomalies will be excluded.
But the question is, 50% of what? 50% of all users? But I alone can turn on many computers, run many programs on each computer, and simultaneously manipulate an infinite number of “users.” Network currency also cannot go around checking one by one whether users are living people; so what is used to define 50%? Bitcoin’s solution is: computing power.
Bitcoin stipulates that all bookkeepers must engage in a computational competition. Each participant must use a particular algorithm to package the ledger entries over a period of time. If two people disagree about the ledger, the packages they produce will of course be different; they do not recognize each other’s versions, while other participants who recognize one package will continue building the next package along that same package, and so on. In the end, the divergence appears as different branches, and the cheater can only keep packaging along the wrong ledger branch. So which branch will a neutral participant choose? The one with the greatest total packaging difficulty. Every package is the result of a computational contest, and we can know under how much difficulty that package was completed, and thus calculate the total difficulty experienced along a given branch.
And the difficulty of the computational contest becomes greater as participants increase—bitcoin is designed so that no matter how much total computing power participants have, the difficulty of the computational contest is always adjusted so that roughly only one package is computed every 10 minutes. That is to say, on the main chain with more participants, the contest becomes harder and harder, so that cheaters with weaker computing power cannot complete an equally strong ledger branch.
In short, the so-called 50% in the bitcoin world depends on the network’s overall computing power. A small group of cheaters cut off from the mainstream—no matter how many virtual accounts they set up, their computing power is still physically limited. To carry out cheating—the so-called “51% attack”—one must deploy massive computing equipment in the real world for it to be possible.
This sounds very difficult. In fact, the total computing power of the entire bitcoin network has already surpassed the combined computing power of the world’s top 500 supercomputers by many times. But the danger of a 51% attack is real, because bitcoin mining has begun to move toward centralization and scale (more on that later). At present, the industry leader, the Lao Mao company (Chinese, of course), can already occupy 20% to 30% of the whole network, and many of the later-emerging cryptocoins besides bitcoin are often, or have already been, subjected to 51% attacks.
But a 51% attack is not as terrifying as imagined. In fact, many later-emerging cryptocoins (generally called altcoins in China, all created by imitating and modifying bitcoin—more on that later) have survived after undergoing 51% attacks. FTC recently experienced a 51% attack and its price rose rather than fell afterward; some people call the 51% attack the “rite of passage” of altcoins, saying that only by withstanding this test does one count as mature. Bitcoin itself has not been free of problems either. Just this year, after one of bitcoin’s software upgrades, old versions did not recognize the new coins generated by the new version, causing a severe split in the bitcoin network. But this problem was quickly patched, and after a brief halving, bitcoin’s exchange rate quickly recovered. If bitcoin’s network truly faced the crisis of a 51% attack, it could always be prevented by upgrading the software, but there is no need for that yet.
In short, mining is the act of competing for bookkeeping rights through computing power, and its significance lies in ensuring the security of bitcoin transactions. But what benefit do participants gain from expending computing power to take part in the competition? Bitcoin’s design is that each time the first person to successfully package a ledger entry gets a reward of newly created currency, plus all the fees contained in that entry. The reward for newly created currency decreases gradually: in the first four years, each packaged block (about every 10 minutes) yields 50 coins, then every four years it is halved, eventually approaching 0. At the same time, however, as bitcoin transactions become increasingly active, the fees generated by transactions may also increase day by day, so competitors participating in mining are always able to get some benefit. If there were no benefit, no one would participate; but if there were no cost, participation would become uncontrollable. The purchase and maintenance of computing equipment, as well as the electricity consumed while the equipment runs, are the costs of participating in mining. Generally speaking, mining revenue should approach mining cost, slightly exceeding it. By the way, recently the emergence of new ASIC miners has greatly reduced mining costs, so bitcoin prices are under considerable downward pressure. But as a large number of new miners are being and will be brought online, mining output will also rapidly become lower (because the total mining reward the world gets every 10 minutes is fixed; more participants means a lower chance of winning), so relatively speaking mining costs will soon soar again (mining cost is relative to output, that is, the cost required to mine one coin; in the long run, this cost should be close to one coin).
What is an ASIC miner? Simply put, it is a machine specially designed to perform bitcoin bookkeeping calculations, and it can do nothing other than mining. But precisely because of this targeted design, its computational efficiency is extremely high. Over the past few months, on the one hand bitcoin prices have soared, and on the other hand the mining-machine revolution has reduced mining costs, so at a certain period mining generated windfall profits. Miners and mining-machine manufacturers reinvested mining proceeds into new miners, rapidly expanding the scale. But investors who obtained windfall profits through mining were limited to that first batch of people daring enough to gamble when specialized miners had just been invented. That period has long since passed. No matter how bold you are now, you can’t make a windfall profit by buying mining machines; it is even almost impossible to recoup your capital. (Of course, you may still make a windfall profit by selling mining machines.) If you really want to gamble, it would be better to play SatoshiDice. Buying mining machines now is not called gambling; it is called giving away money.
The reward that the entire bitcoin network gives all miners over a certain period is basically fixed: at present it is 25 coins every 10 minutes, that is, 150 coins per hour. So if 100 identical mining machines participate, each machine can on average get 1.5 coins; if 10,000 machines participate, that becomes 0.015. As the number of participants increases and more miners are brought online, the reward that one unit of computing power can obtain becomes smaller and smaller. Moreover, under current conditions, and under foreseeable future conditions, the difficulty of mining will grow exponentially. It doubled in the past two months or so, which means income was halved. If this trend continues (in fact, the growth rate in the second half of the year is more likely to accelerate), then even now spending 1 coin to buy 1G/s of computing power can hardly recoup costs, let alone the futures miners that many manufacturers are offering for August and even September, which are also priced above this level.
Why can’t one recoup costs? The reason is simple: the sum of a decreasing geometric series converges. 1 + 1/2 + 1/4 + 1/8 + 1/16 + …… = 2. This is mathematical common knowledge that one should know from middle school, but many people forget it, or never think about it at all. That is to say, if mining output is halved every two months, then the total number of coins that a certain fixed amount of computing power can mine over all time is roughly equal to twice the output from the first two months of mining, or the output from the two months before mining began with the same computing power.
Of course, this does not take into account the possibility that the growth of real bitcoin network computing power may stagnate, but it also does not take into account further accelerated explosive growth of computing power, or miner failures, miner vendors defaulting on delivery, and so on. Some people say that because I am optimistic about bitcoin, I buy mining machines. This is the wrong way around. The better bitcoin develops, and the more participants there are, the harder it becomes for mining machines to recoup their cost. Only by hoping that the bitcoin network falls into stagnation can mining machines recoup their cost. Others say: although bitcoin output has fallen, if bitcoin appreciates in the future, then wouldn’t I also recoup my cost? That is the illusion I mentioned at the beginning—being misled by fiat currency. Spend 2 coins now to buy a mining machine. The exchange rate is 500 yuan per coin. In the end you mine 1 coin, and it rises to 10,000 yuan. Wouldn’t that be making a huge profit? This line of thought is ridiculous: putting in 2 coins and getting 1 coin out, and yet calling that a profit. Buy 2 coins now, and a year later, two years later, they are still 2 coins. 2 > 1—this is probably common knowledge even for an infant, but for many mining-machine investors it has somehow been lost.
The mining-machine market in China has developed explosively. Many teams, reliable or not, have gotten into the mining-machine business. Lao Mao buys spot mining machines at jaw-droppingly high prices, yet he is still, after all, a relatively trustworthy company—after all, the price is set by the market; if the seller charges a high price and still sells out, what can I do about that? Should I expect merchants to develop a charitable impulse and consciously earn a little less? The ones still doing futures mining machines now are generally unreliable. But what is truly absurd is computing-power rental—I no longer sell mining machines, but instead claim to be deploying data centers, letting customers buy already deployed computing power, for example: how much does it cost to rent 1G of computing power for one year? Then consumers are spared the procedures of machine allocation, management, maintenance, and so on, and even electricity is included for you; the corresponding mining output is simply paid to you every week. What a fine scheme—practically tailor-made for Chinese newbies. Some of the earlier services of this sort still advertised themselves as “mining-machine leasing,” and also required you to surrender about 8% of the output. More recent companies directly sell “computing power” and do not require any commission, and even provide a service to help you convert it into renminbi for payment. You just need to hand over the money and sit back to enjoy the proceeds. Where in the world is there such a good deal?
Theoretically speaking, with the same computing power, buying a mining machine and maintaining it yourself is obviously inferior to buying computing power and letting someone else handle everything for you. But where is the problem? The problem is that the merchant doesn’t need to deploy even a single mining machine at all; they can simply use the customer’s money to pay the customer back. Obviously, your investment will never recoup its cost. You put in 2 coins, and this week I give you 0.1X coin. You think, great, 20 weeks to break even. In the later weeks the income drops a bit, but in two months it’s already more than 0.5 coin, right? But after a few more months it may only be 0.0X coin each time, and even after more than half a year it may be only 0.00X coin. In the end, over an entire year you may receive only 1 coin in total, and the rest is the merchant’s effortless net profit…… If you really want to buy, I can provide this service too: give me 13 coins and I’ll sell you 10G of computing power for one year~
The reason I’m saying so much about mining-machine scams is that I truly feel pain over the fact that so many Chinese people are putting money into mining machines. Of course, among them there are some speculators who richly deserve to be cheated to death, but there are also quite a few participants who are actually motivated by good intentions. On the one hand they are optimistic about bitcoin, and on the other hand they hope to make a long-term investment with steady, trickling returns. But that is precisely the mistake. The proper way to be optimistic about bitcoin and invest long-term is to hoard coins; those who are not optimistic about bitcoin, or who want to speculate in the short term, are the ones who should buy mining machines. Many people who buy mining machines do not realize that they are making a choice contrary to their own intentions. This is hardly surprising, because in an inflationary fiat-currency system, no one regards holding currency as a long-term investment, and the returns of a “money printer” are of course long-term. People would rather buy a hen that lays eggs than a basket of eggs. The original intention behind this way of thinking is good, but in the case of bitcoin, our mindset has to be adjusted.
Painful as it is, for many new entrants and newbie investors, I also don’t have much sympathy. Earlier I saw many transaction records and positive reviews in some Taobao store selling computing power, and I was shocked to see one reviewer say: “The boss is really nice, patiently teaching me how to install the bitcoin client……” @#$#^%#$…… Brother, you haven’t even figured out how to install the bitcoin client, and you’ve already rushed over to spend thousands or even tens of thousands of yuan on a mining machine? So when it comes to people like this, Mandarin really lacks a proper adjective—“newbie” only refers to the ignorant who know nothing of the matter, but for someone who is not only ignorant yet especially fearless, charging at the very front line of the slaughter, I can only describe them in Shanghai dialect: sengtou (chongtou, congtou). People like this are simply asking to be fleeced.
II. Altcoins: Exploring Different Mining Mechanisms
We have already explained that “mining” is the operating mechanism by which the Bitcoin network is sustained, and by now Bitcoin mining has entered an era of specialization. Efforts to improve Bitcoin have also been ongoing all along. Of course, one approach is through the efforts of foundations and open-source communities, maintaining updates to Bitcoin itself—by now Bitcoin has been released up to 0.8.3, and many major and minor improvements have long since been made. At the same time, others directly modify Bitcoin’s source code and release one new currency after another. These new currencies are also based on the Bitcoin idea: a decentralized distributed network structure, sustaining the entire transaction system by having the whole network jointly verify and record every transaction, while the participants who maintain the network have the chance to receive new coins and transaction fees as rewards—most new currencies have followed this model, merely making different adjustments in certain respects.
Thus these new coins are called “shānzhài coins” by Chinese people; foreigners, of course, do not have such a handy term. They call these coins “Alternate cryptocurrencies,” which literally means “alternative cryptographic currencies.” Considering the pejorative sense of shānzhài, “cryptocoins” as a neutral term may be a bit more appropriate. But in fact the vast majority of shānzhài coins do not even deserve to be called shānzhài; they can only be called “clone coins.” Typical examples are our Chinese masterpieces: China Coin (CNC), Yuanbao Coin (YBC), Dragon Coin, Cosmos Coin, among many oddities that I won’t introduce in detail. These coins simply take the source code of Bitcoin, or of coins improved by earlier people, and change a few parameters, such as how many minutes it takes to generate a block, how many coins are rewarded to miners for each block, the total supply and issuance speed of the coin, and so on. Some even go online with not a single character altered, merely under a different name. Among them, some developers are just having fun out of a spirit of amusement, but others—especially the large number of new coins that have recently emerged in China—are, to put it simply, all about cheating people out of money. There are too many greedy rubes in China, and plenty abroad as well.
A lot of people think this way: I missed the chance when Bitcoin had just been invented; back then mining must have been hugely profitable, right? Then now I should also get to start from the “original shares.” We can look at the slogan of “Yuanbao Coin,” which very typically captures this mentality: “Bitcoin has, in just two short years, taken the world by storm, and is gradually destroying the old financial order while building a new one. Those who first learned of and believed in Bitcoin are now rich beyond measure. However, Bitcoin now faces challenges unforeseen by its founders. With the emergence of ASIC professional mining machines, mining has become increasingly difficult for ordinary users, to the point where even electricity costs can no longer be recovered, and the momentum for Bitcoin’s further spread has gradually been lost. Bitcoin’s early large holders were mostly concentrated overseas; Chinese users, because of the language barrier, have already missed Bitcoin’s earliest and fastest period of appreciation, and can only stand at the top and take the bag; the endless stream of shānzhài coins is dazzling, and one can even be tricked into losing everything if not careful…”
This reason really does sound like it makes a bit of sense. Back then Bitcoin was published in English on foreign forums, and we didn’t see it, so we missed it—that’s not our fault, that’s not fair. Let’s start over now and have the Chinese people build our own system~ But is that really unfair? In fact, Bitcoin remained a niche phenomenon in English-speaking countries for a long time too, while introductions in Chinese circles had already appeared several years earlier; people who now rank among the world’s top holders of Bitcoin, perhaps even number one or two, are also in China. Russia and other non-English-speaking countries were also relatively ahead in Bitcoin’s early development. The emergence of any new thing will always be encountered by some people before others; that is not a problem at all. Moreover, encountering it earlier is not necessarily better. In fact, many early Bitcoin players mined a few blocks and then, thinking it was boring, uninstalled the software, and those coins were lost forever. Others, seeing Bitcoin go from worthless to worth a bit of money and then even rise dozens of times over, had already sold out. It was actually those who entered a bit later—Li Xiaolai and the like—who made the big money.
Of course, if you say that in our vast China we should also make a coin of our own just to have some fun, I support that too. But what is the point of producing something that is a complete copy, merely to make a fool of oneself? Then either the market is manipulated and the money is pocketed before running off, or at best one merely adds another piece of garbage to the already rampant lineage of shānzhài coins. Wouldn’t that just bring shame on Chinese people? The one branded as “China Coin” has already cheated plenty of people and is well on its way to taking the crown as the ultimate scam coin; I wonder whether Yuanbao Coin can overtake it from behind~
Below I will only discuss a few coins that, beyond merely tweaking parameters, also include certain substantive changes:
1. Namecoin — the Byproduct Model
Namecoin (NMC) may have been one of the earliest shānzhài coins. It chose to attach itself to the Bitcoin network and adopted a “merged mining” model, meaning one could mine Bitcoin while simultaneously mining some Namecoin. The earliest few shānzhài coins all adopted a merged model, because at the time Bitcoin was still in an earlier stage, mining had not yet shown obvious windfall profits, and there was not much competition. People interested in cryptocoins would naturally prefer to mine Bitcoin first, and there was no obvious reason for them to give up Bitcoin and mine another coin instead. So the merged mining model—positioning itself as a “byproduct” of Bitcoin—was a relatively desirable solution.
What is usually translated as “Namecoin” is in fact often rendered as “name coin,” but here the “name” refers to domain names, so for the sake of clarity and fluency, I think “domain coin” is a better translation. This hints that one of the major creative ideas of this coin is that it further provides a “byproduct” of its own—a decentralized domain name system.
We say the Internet is decentralized, but in fact certain aspects of the Internet still retain some undeniable “centers,” such as the domain name system. If we want to register a domain name, we are forced to turn to a domain provider; at the most fundamental level, all sorts of domains are allocated under the auspices of ICANN, an international organization. Although it is nominally independent of the U.S. government, in fact the domains of every country must obey the management of a central authority, and ICANN itself can hardly guarantee fairness and neutrality. In any case, the allocation of domain names remains a top-down hierarchical structure rather than a decentralized network structure.
Inspired by the Bitcoin idea, Namecoin uses a Bitcoin-designed model to maintain a domain name system: every participant keeps records of ownership and pointing for all domains, and every new domain registration or change is also broadcast to the whole network; changes that receive 51% approval are recorded. This 51% is still based on computing power, just as with Bitcoin. The workload of participating in verification has the chance to receive a certain amount of Namecoin as a reward, while when I register or renew a domain, I have to consume the corresponding Namecoin.
This mechanism very ingeniously solves the problem of domain name decentralization. Traditionally, registering a domain always means registering with some institution, whereas here we do not register with a center, but with the entire network. Of course, registration is not arbitrary; it requires consuming Namecoin, and this Namecoin is the reward for participating in maintaining the network, or of course it can also be purchased from other users with other currencies. Namecoin now supports .bit domains, though such domains do not belong to the general Internet domain system and require special means—through the Namecoin network or proxy servers—to access.
This mechanism is marvelous; the only problem is that it was invented too late. The mainstream domain allocation system had already developed to a sufficiently full and widespread degree, so that it is hard to detach from it, and without any special reason there is no need to go to the trouble of using an additional system. So domains in the Namecoin network are still rarely used today; I myself have registered philo.bit and 510.bit. Recent client updates have greatly facilitated domain registration and may attract some new registrations, but this domain system still shows no prospect of popularization. Unless ICANN were exposed in some unimaginable scandal, people will still trust its services.
In addition, because Namecoin is “useful,” its value probably will not surge very much, because it should always remain within a range that makes domain costs acceptable. Of course, open-source software developers can adjust domain prices when upgrading the software, but that will eventually lead to disagreements. So I am not very optimistic about Namecoin’s prospects for appreciation. But as the first popular shānzhài coin, and one that is richly creative, Namecoin is worthy of respect.
The Namecoin system also reminds us that byproducts derived from the Bitcoin idea need not always be currencies; they can also be various other decentralized network systems. For example, the recently emerged decentralized instant messaging system Bitcoin Message is one case: this system ensures that communications neither pass through nor are stored on any single central server, and will not be intercepted or eavesdropped on.
2. Litecoin — Resisting the Specialization of Mining
However, the one now sitting at the top of the altcoin heap, or to say the second place among cryptocoins, is undoubtedly Litecoin (LTC). As its name suggests, Litecoin claims to be a lightweight version of Bitcoin; relative to Bitcoin, its block generation speed is faster, and the time required to confirm transactions is correspondingly shorter. Litecoin’s total supply is four times that of Bitcoin—basically these are just simple parameter changes. After Litecoin, one can simply swap out the values for block generation intervals and total currency supply to make yet another new coin.
The biggest difference is that Litecoin and Bitcoin use different cryptographic algorithms: Bitcoin uses SHA256, while Litecoin uses scrypt. In fact, it is hard to say that one of these algorithms is more reliable or more secure than the other; the main difference is that the latter requires a large amount of memory during computation.
By the way, Litecoin was not the first shānzhài coin to use the scrypt algorithm; the first was tenebrix, but its developer was rather unscrupulous and mined several million coins for himself before public release. Such behavior, in an open-source system, of course cannot be concealed, and it was scorned in the open-source community. Then the author of Litecoin simply copied tenebrix’s source code and reissued it under the name fairbrix, with the only difference being that he did not pre-mine it. But this coin was immediately hit by a 51% attack after launch and was eventually abandoned. After painful reflection, the developer learned his lesson and this time did enough pre-launch warming-up in the open-source community before releasing it—that was Litecoin, which in fact was already the third scrypt shānzhài coin.
We find that, in fact, Litecoin’s technical content is not especially high. Its main change was simply to alter the algorithm, and even that change was directly copied from elsewhere. Of course, after it became increasingly popular, the developer team still did a very active job maintaining it, but overall its spirit of innovation was not particularly strong. In my view, Litecoin’s popularity was mainly a matter of timing and circumstance.
The adoption of a new algorithm was, at the time, a response to the emergence of GPU (graphics card) mining. Bitcoin originally used only CPUs for its computing-power competition, and later someone discovered that by using GPUs—especially ATI graphics cards—the efficiency could be increased by more than a hundredfold. This ushered in the first “arms race” in Bitcoin mining history. Mining graphics cards suddenly became highly sought after, and some people began deliberately purchasing large numbers of graphics cards to set up powerful mining rigs, while mining with CPUs alone could basically be ignored.
Under the scrypt algorithm, however, although GPUs still have higher efficiency, CPU output is not so negligible as to be ignored. A group of old miners who had already been eliminated by the technological revolution found a new outlet, and Litecoin can be said to have arisen in response to the times.
More recently, with the appearance of Bitcoin ASIC professional mining machines, the scrypt algorithm has once again become fashionable. Altcoins newly released this year no longer use the SHA256 algorithm. Because ASIC miners are tailor-made for mining Bitcoin and can do nothing but compute SHA256, they naturally cannot mine scrypt altcoins. Thus, when ASIC miners once again left graphics cards far behind in efficiency, the second batch of old miners eliminated by the technological revolution also threw themselves into the arms of scrypt.
So we can understand the reason scrypt has been favored: in short, it is resistance to the “specialization” of mining. This mentality is understandable, but I do not agree with treating it as a special advantage of Litecoin. In fact, no matter which algorithm it is, it will sooner or later tend toward specialization. Just as in any sport or game competition, if it develops sufficiently well and the prize money is high enough, sooner or later there will be professional athletes and professional teams, and professionally trained athletes will certainly leave amateur competitors far behind in competitiveness. To avoid the emergence of professional athletes, either one must keep changing the rules of the competition, or one must keep the competition from ever becoming too hot. In other words, the emergence of specialization is precisely a sign that Bitcoin has become hot enough and is maturing. The fact that it is currently difficult to manufacture professional miners for scrypt is apparently only due to the high price of memory components; if Litecoin continues to heat up and appreciates dramatically, memory will by no means pose an obstacle to manufacturing professional mining machines. The mining mechanism of cryptographic currencies that copy Bitcoin’s system basically all has fixed output over a certain period of time, which means that as participants increase, the earnings of any single participant will inevitably be diluted, profits will become negligible, and in the end specialized, large-scale miners will become the main force maintaining the network.
This is not some sort of departure from Bitcoin’s original intention. Decentralization does not mean flattening all hierarchical structures and role divisions; just as the Internet does not require every computer to be a server. Most ordinary people are merely users of the network and do not need to bear the responsibility of maintaining it. As Bitcoin develops, most people should first and foremost be users of Bitcoin, not contributors.
Ordinary users, especially those with outdated equipment, being eliminated from Bitcoin mining is a normal trend. This marks Bitcoin’s transition from a novelty toy among a small circle of early enthusiasts to a stage of popularization and practical use. Early participants supported Bitcoin mainly by mining, whereas participants today, if they want to support Bitcoin, should find the best way to do so by using it and expanding its circulation. Clinging nostalgically to the era of全民挖矿 is merely a sign of immaturity.
I do not completely deny the significance of new algorithms. In fact, once the Bitcoin network has already become sufficiently mature, newly born altcoins may have to adopt new algorithms in order to ensure they are less vulnerable to 51% attacks and more attractive to old players, so scrypt has almost become a standard feature of recent altcoins. But it must be pointed out that this kind of innovation is not innovation in essence; rather, it merely catered to the nostalgia of some old players during the Bitcoin mining revolution. In short, it is a product of a transitional period. Of course, Litecoin caught the right time and conditions, and achieved enormous success; owing to inertia, its position will likely continue for some time. But from the philosophical standpoint of someone who cares more about ideas than profits, Litecoin is not very interesting.
I am now also using graphics cards to mine scrypt-based altcoins and then immediately converting them into Bitcoin; at present this is a bit more profitable than mining Bitcoin directly. New multi-coin auto-pools have recently appeared, such as https://www.multipool.in/, which automatically monitors the market exchange rates of all altcoins (in trading against Bitcoin) and the difficulty of mining across the entire network, and automatically chooses the coin with the highest output value to mine. Another mining pool, http://middlecoin.com/, can also automatically convert the payout into Bitcoin before sending it. Friends who have good graphics cards can still mine a bit, but of course one should not buy a graphics card especially for mining.
3. Peercoin — The Equity Model, Energy Saving, and Inflation
PPcoin (PPC), Peer-to-Peer, peer-to-peer coin. In fact, Bitcoin is obviously also a “P2P” currency, so the author Sunny King did not really give this coin a distinctive name. But Sunny King is undoubtedly the most serious of all shānzhài coin authors, and he has recently developed a prime coin that takes an approach different from both Bitcoin and PPcoin; I will talk about that later.
Just like Bitcoin’s creator Satoshi Nakamoto, Sunny King always first provides a white paper, or rather a paper, for the new coins he develops. This is very important, because it means that this new currency is first of all an idea; its open-source program is the practice of a theoretical exploration, rather than merely a copy or a follow-the-crowd imitation. By contrast, most other shānzhài coins are nothing more than tweaks to parameters on the basis of predecessors, and simply cannot write any paper at all; at most they can provide a wiki introduction page listing the various parameters, without any way to demonstrate where the significance of this new coin lies, or what exactly it is exploring. Whatever one may think of Sunny King as a person, just for these two papers alone, I respect him, and believe he is on a different level from those shānzhài coin developers who exist only to fleece money.
Peercoin’s main innovation is the entirely new “Proof of Stake” model. Compared with Bitcoin’s “Proof of Work,” Peercoin uses a PoS+PoS model to confirm transactions. As I have already explained, Bitcoin does not rely on a centralized institution; instead, it relies on the whole network’s 51% to confirm the legitimacy of transactions. And this so-called 51% is determined in Bitcoin by computing power: miners take part in a computing-power contest, and in the end “work” is the yardstick by which rewards are measured. In Peercoin’s case, the initial stage still mainly relies on Proof of Work, but as the amount of Peercoin held in the network gradually increases, it will progressively shift to being based mainly on Proof of Stake—that is, the right to speak is determined by how much currency a participant holds. In PoS mode, the 51% is determined not by hash power but by the amount of coins held. And the remuneration for participating in maintaining the transaction network through PoS is also distributed according to the proportion of coins held, roughly equivalent to an annual interest rate of 1%.
This gives rise to several “benefits.” First, the cost of a 51% attack becomes higher: one must not only control more than half of the network’s computing power, but also possess more than half of the currency supply. As time goes on, this will become increasingly difficult to achieve.
Second, as the total amount of coins held increases, the returns from the computing-power race will diminish, so in the long run the mining industry’s “arms race” will not continue endlessly; therefore maintaining the Peercoin network will be very “energy-saving.”
Third, an annual interest rate of around 1% means that Peercoin’s total supply does not have a fixed upper limit, yet it will not be arbitrarily overissued by some central authority; inflation is slight and predictable. (In fact, gold also increases by about 1% a year, but basically we still regard it as deflationary.)
The reason I put “benefits” in quotation marks is that so-called benefits are always relative. I do not think the ideas behind these Peercoin improvements are ineffective or futile, but it is not necessarily the case that these improvements make Peercoin superior to Bitcoin.
In the previous article I already mentioned that I do not think Bitcoin alone can rule the world of money. Decentralization is a characteristic of the network age, but diversity is too.
If the money system of the future is an ecosystem, then only currencies occupying the same ecological niche will engage in a life-or-death struggle for survival, whereas currencies with different modes, different positioning, or even different values will occupy different ecological niches. It is these different species that may coexist and prosper together. And at different positions, the standards for judging good and bad are also different. For example, a cheetah is of course faster than a sloth, but that does not prove that the cheetah species is superior to the sloth species. Running fast is an advantage for a cheetah, but placed on a sloth, it may not be an advantage at all.
Even in the age of gold, silver and copper coins also long coexisted with it as supplements. Gold may not even be the best element for serving as the main currency; gold’s status also had all sorts of accidental factors of timing, place, and human support. But even if a more superior metal were discovered in the future, there would be no need to replace gold’s status.
In fact, no species is perfect, and it is also hard to say that there is one metal that is the most perfect for use as money. For example, although gold is excellent in every respect, it is too scarce, so its mining is too easy to monopolize, and it is not very convenient for everyday small transactions. Copper coins, by contrast, are produced in larger quantities and more dispersedly, but they are also prone to flooding the market and to forgery. Silver is relatively more balanced. Such coexistence of multiple currencies, each with its own strengths and weaknesses, balancing one another or alternately gaining and losing ground, is what constitutes a complete monetary ecosystem.
So, we may regard Peercoin’s energy efficiency, slight inflation, and so on as advantages, but that does not mean that Bitcoin must necessarily have corresponding “disadvantages.”
So does Bitcoin really have serious flaws? Of course Bitcoin is not perfect, but some problems may not be things that must be corrected.
First, let us talk about the possibility of hash-power monopolization brought about by hash-power competition. This is indeed real, but under the spontaneous regulation of economics and the entire network’s maintainers, it is not easy to achieve. And the ever-present sense of crisis sometimes also plays a positive and promoting role.
Second is the energy-consumption issue. One must admit this is a major problem. In the long run, Bitcoin’s cost should approach the output of its mining, and once Bitcoin’s value rises further, its mining costs will also skyrocket without limit. But the cost of each mining device itself and its electricity consumption are fixed. In the future, more identical devices will have to be expended to mine the same amount, which means that the energy consumed to mine the same amount will keep rising.
This mechanism is not without drawbacks. Mining input and output reach a certain balance according to market laws. As the production of new Bitcoins decreases and the currency value stabilizes, the mining industry will not always expand explosively as it has recently. On the one hand, mining input will still keep increasing, but not by an absurd margin; with technological progress reducing the energy consumption of computing equipment, total energy use will not lose control either. It should be said that in the foreseeable future, maintaining the Bitcoin transaction network will always be far more energy-saving and environmentally friendly than maintaining any country’s fiat-currency system.
But what about the future? What if the competitor is not a fiat system but another cryptocurrency? Then, on the single point of “energy saving,” Bitcoin is indeed a weakness. It is like a sloth certainly being more energy-efficient than a cheetah; the key is that it adopts a different mode of survival. For Peercoin to achieve energy efficiency, it must use Proof of Stake and the annual-interest-rate model, and these modes are not necessarily superior to Bitcoin’s.
What about the feature of slight inflation? Since the new coins of Peercoin’s inflation are obtained entirely through a mechanism similar to “deposits,” rather than first through ex nihilo lending, this inflation is unlikely to run amok. By contrast, Bitcoin’s total supply tends toward a fixed value; because wallets can be lost, in fact Bitcoin’s total supply will, in the long run, keep declining. Of course, whether the total supply is constant or declining does not affect the currency’s usefulness. Even if only one hundred Bitcoins were left in the world, digital currency is extremely divisible, so it would still circulate just fine.
But in the long run, a constantly decreasing total supply will indeed create certain problems. Suppose Bitcoin can remain in circulation for several hundred years; after several hundred years, perhaps most of the coins will have been lost. Of course, whether 50% or 10% remains, the monetary system can still function normally. But perhaps suddenly an archaeological discovery will unearth a large wallet key from among old papers—for instance, Li Xiaolai is said to have paper wallets amounting to nearly 1% of the total supply. And today’s 1% may become 10% eight hundred years later. The result would be a sudden 10% inflation in the year of the archaeological discovery.
In this way, as time passes, Bitcoin will become increasingly worrying, and in fact people do not know whether those currencies unused for decades are being deliberately hoarded, temporarily forgotten, or permanently lost. This makes people’s estimates of Bitcoin’s total supply increasingly unreliable. In this respect Peercoin has no problem, because its 1% annual interest rate is in fact only obtainable by keeping the client constantly on and participating in network maintenance; idle and abandoned coins do not automatically grow.
Of course, this is really being overly worried about nothing. The rhythm of the internet age is measured in spans of ten years; a hundred years or even a thousand years later, the era will already have changed countless times. By then—and perhaps even a few decades from now—due to the development of quantum computing, the entire cryptography-based digital-currency world will be completely overturned wholesale (including the traditional financial world as well). If it survives, Bitcoin will also undergo countless revisions and improvements. Talking now about defects on a hundred-year timescale has no great practical significance. Still, since it does have theoretical significance, the exploration and experimentation of Peercoin are of course very meaningful.
4. Primecoin — mining with scientific value
Primecoin (XPM) is a new coin Sunny King released recently. I happened to catch its release, and on the very first day after it came out I joined the mining, mining 300 coins. Later, this coin received enormous attention; those 300 coins could for a time be sold for more than 5 Bitcoins (almost three thousand yuan), and now they can also sell for about 2 Bitcoins. Unfortunately, I sold more than 100 coins at a slightly lower price and kept the remaining nearly 200 coins as a collector’s hold. Clearly, I picked up a bargain for nothing. And this was not merely luck; I had already been optimistic about this coin’s prospects beforehand.
This is not to say that I really fully agree with the significance of Primecoin. First, I believe that whether or not Primecoin solves a flaw in Bitcoin, at the very least many people do in fact regard it as a flaw, just as many people regard mining-machine specialization as a flaw. Since there are supporters, it will be valuable.
Primecoin still addresses Bitcoin mining’s “energy consumption” problem; the solution is not to reduce energy consumption, but to try to make energy consumption meaningful.
In fact, Bitcoin is one of the world’s largest distributed computing projects (there should be no “one of” about it, right?). The network’s total hash power has long exceeded several times the combined power of the world’s top 500 supercomputers. Isn’t it too wasteful to have so much computing power merely to maintain the Bitcoin network itself? Other distributed computing projects calculate protein folding, search for gravitational waves, look for extraterrestrials, and do all kinds of scientific work. So if the Bitcoin network has such enormous computing power, how wonderful it would be if it could do some scientific computing!
But one cannot have both the fish and the bear’s paw. Originally, mining was for maintaining the security and reliability of transactions, using the random computation of passwords to compete in hash power. But if the contest’s task is not random decryption but a specific scientific problem, then cheating may be more likely to occur.
Still, is it possible to obtain some scientific effect while ensuring security? I think it is still possible. One method could be a byproduct model like Namecoin’s: since one can mine jointly, doing some Namecoin calculations unrelated to Bitcoin while mining Bitcoin, then of course it is also possible to add other computations. But this does not require developing another currency; it can be implemented by developing an additional application on top of Bitcoin, and then having some mining pool host the scientific project. This is entirely possible. Even if it were only a tiny fraction of Bitcoin’s hash power, it would already be enough to make many other distributed computing projects envious. (Of course, the current problem is that specialized mining machines can only perform SHA256 computations, so perhaps the scrypt-family currencies still mined on ordinary computers, such as Litecoin, are more suitable for doing something like this.)
Another, more thorough approach is to develop a new coin specifically for this purpose. Searching for proteins and the like seems to require concrete research projects designed and arranged by specialized institutions, which does not seem very suitable for a decentralized model; mathematical problems, on the other hand, seem more suitable.
As its name suggests, Primecoin aims to find large prime numbers. At the same time, because of the randomness and unpredictability of the distribution of primes, yet with an estimable distribution probability, using the search for primes to prove work seems like a feasible scheme.
Of course, Primecoin’s own technical design is not necessarily perfect. For example, its method of finding primes seems incomplete, with a certain degree of exception, and mathematically it does not seem to be the best prime algorithm either. I do not know the specific principles, but I guess many compromises were made in order to achieve a balance between the two originally unrelated tasks of finding primes and verifying currency.
But I think there is no need to be too harsh too early on Primecoin’s imperfections. After all, first and foremost it is a paper, a new idea; only secondarily is it an experiment carried out along that idea. Despite all the compromises, it did indeed graft network currency and scientific computing together. On this basis, we can look forward to its further development.
For example, the aforementioned additional application that provides some scientific computing as a byproduct is more proper and legitimate when attached to Primecoin than when attached to Bitcoin or Litecoin. I even imagine that this “extra scientific computation” could be embedded within Primecoin itself: any participant on the network can publish a computational task, and as long as the task can be distributedly computed according to certain rules and its difficulty is estimable, researchers can publish the problem and a bounty; miners decide whether to participate in the computation based on the return rate of the bounty and the difficulty, and miners who participate in the computation receive the prize paid by the bounty poster according to their work. When no one is offering a bounty, miners continue to process prime-number problems. And the answer to the problem can either be made public to the whole network, or be solved in encrypted form, viewable only with the bounty poster’s own private key. In this way, a truly distributed distributed-computing system becomes possible. Of course, this design probably still has many technical problems, but Primecoin is the first step in opening up the line of thought.
bitcoin:1411wgvec6AKWa5cp7sYqNdo8bq6T4xpe6
Translated from the Chinese original with AI assistance. The original text is authoritative.
Leave a Reply