Viewpoint | Nothing is cheaper than PoW, Part-2

Viewpoint | Nothing is cheaper than PoW, Part-1

Give two examples: TENDERMINT and DPoS

First, let's take a look at the incomprehensible consensus algorithm that transforms PoW.

Example 1: Tendermint/Casper “… In addition, your profitability depends on the behavior and usability of other pledged verifiers…”


Tendermint is such a solution: an individual can be a member of the “verifier” collection by locking a portion of the electronic cash as a deposit. If the verifier does evil, it will lose this part of the “deposit”; at the same time, as long as enough good verifiers exist, the entire network can reach a consensus by “zero mining”.

In fact, personally, this is exactly the same as Bitco's PoW.

Put aside these first…

This is a blog about economics. I will ignore all (more fundamental) technical issues about Tendermint (eg, too much, signature verification scalability, certifier offline/ddos), operational issues (eg , how the value of the contemporary currency is too low to start the network, how to control when the price of the token falls, how to create a marginal benefit higher than the bitcoin), and the problems of the cryptography system (for example, using a cheap personal cloud disk can launch long-range attacks, Create millions of pseudo-chains that look like main chains [preparing for large-scale double-flower attacks] and also crowd out nodes that try to point to the correct main chain). The above questions are not discussed in this article, and I will assume that the system is running normally, and there is no squad or seduce between the verifiers.

Verifier vs. miner

The problem now is that the verifier's marginal benefit MR (outside bonus + transaction fee) will also affect the marginal cost MC. In the existing bitcoin system, there is a consensus that every 10 minutes the miners “wasted” bitcoin worth about 25.01 USD; if Tendermint agrees, every ten minutes the verifier will waste bitcoin worth about 25.01 USD.

Unproductive input = "waste"

What is the difference between the following two columns?

01 In the explicit PoW example, the miner secures the network. For example, the whole society only produces quantitative good cars, brand-name bags, and luxury houses (bitcoins), and miners can get some of these scarce resources every year (other workers do not get these resources). These people turned to the upper reaches of the miners (mining hardware suppliers or power plants), while others turned to the lower reaches of the miners (miners favorite restaurants, charities, artists, or family and friends).

In the case of the implicit PoW, those who have a verification deposit will get a proportional benefit; they can drive a luxury car, buy luxury goods for their wives, and so on.

I believe that "liquidity is very important" is it so difficult?

I am very convinced that the economic concept of "liquidity" is not a popular word. But the simple understanding is that liquidity represents a quick and stable exchange of money. Isn't this like the emergency reserve we store at home?

It is a pity that the infinite zero-interest rate policy of today's society ruined a whole generation of potential savings opportunities (cannot become "indirect investors"); you can ask your parents how they can buy a wealth management product called CD. Lock the money for a period of time (6 months, 1 year, 2 years); the inconvenience caused by the lock of the money, correspondingly bring higher interest (because the bank can lend at a higher loan rate ). You can use this to “sell” excess liquidity and let the entire society allocate liquidity more efficiently.

You can ask Lehman Brothers or the victims of Zerg Ruch to know how important "liquidity" is. If the liquidity of the funds is all right, these people's carefully calculated long-term plans will always benefit them; but things are not so simple.

For those who are completely unwilling to share their mobility (even if they buy a CD, early redemption may occur – as a penalty, the holder of the early redemption deposit in Tendermint cannot take out all the deposits within one year) The whole society has "nothing" from the economic level, and these non-current assets, the potential liquidity that cannot be transferred, is a kind of "waste."


Make a wishful calculation

Some people may ask, "Isn't the lock-in time of the deposit cancelled in some way?", "Is the whole society really damaged because some money supply is locked out and cannot be circulated?"

The answer is: No, the lock time will not be canceled, and it will not cause damage. All PoWs (explicit and implicit) will have a positive impact, because the bits on the miners only work under the normal operation of the Bitcoin network. The currency is valuable. The benefits of this include the fact that miners make money, chipmakers have orders, companies that provide services are profitable, inefficient payment systems are upgraded or replaced, and so on.

Statement 1, 3, 4

In this Vitalik blog, VB acknowledges that “one of the objections to the long-term pledge of funds is that this will lead users to lock in funds; this is an inefficient behavior, the same as the mining loss caused by PoW” . VB also lists four "response explanations" . There are first and third points that are completely wrong with the horse's mouth: [Declaration 1] The marginal cost is not the total cost (VB is obviously measuring the marginal cost of a single block, not the marginal cost of the work unit), [Declaration 3] deposit It's absolutely safe (it's just like saying "mining hardware is working very efficiently").

Finally, [Declaration 4] The long-term pledge allows us to directly punish the (funds) “reserves” of the verifier, rather than only affecting its (operating) “cash flow”. This has also been confirmed, because bitcoin mining has been dominated by ASICs and well-known mines (the economic value of the mines will be completely destroyed in the attack) (Translator's Note: mining activities) There is also a subject's capital value that will be directly affected).

Statement 2

[Statement 2] Capital lock-in will result in price deflation that is beneficial to the entire system. This is completely wishful thinking. “Reducing the money supply for trading purposes” is unreliable; and such statements distorted the meaning of the deposit, because the only reason for the certifier to freeze their funds is that they can The money is used out . If you really want to tighten the supply, you should permanently destroy the money. This destruction is at the expense of the producer's sacrifice to maintain a high level of life for non-producers. In contrast, bondholders can trade in today's relatively higher levels of life simply by sacrificing today's enjoyment.

Monetary economics is probably the most confusing branch of economics, so I use three other ways to explain it.

Statement 2(i) – You have to choose a runway (you have to choose a definition)

First, the “price level” will only be affected by changes in the money supply (and, if the currency's derivative market is active and well maintained, only the unpredictable changes in the money supply can produce this effect). What is said here is that even if the capital lock changes the money supply (although it does not), as long as the lock amount is constant, the deflation effect will not appear. If the supply changes, the beneficial deflation brought about by it will be offset by inflation.

It should be noted that in the short period of time when the funds are frozen, these currencies are not used for “circulation”… If the currency is not circulated, it can bring favorable price tightening, then everyone will tend to put money in hand. Freeze (not just the verifier) ​​(Translator's Note: If there is such an effect, then there is no need to design).

Therefore, we either consider the expected flow in the future (including the deposit), or we will ignore it all (only the current maximum supply of money).

Statement 2(ii) – Defence Act

Second, assume that the effect of freezing the currency and destroying the currency on the price level is the same. If 99.9% of the money supply is locked, the price of the commodity should fall to 1/1000 (everyone knows what the liquidity will reduce by 99.9%). Now, a small number of people who still have cash ("liquidity") can spend it madly. They can even join forces with "the big crocodile who has a huge amount of money in their hands and lack cash flow" to invest in other new projects. ", plunder a lot of good things. This kind of “project” will make more money than any non-financial project (for commercial activities, factory opening, car loans, etc.), and then the rising interest rate (the “liquidity” price) will cause the industrial project to be squeezed out of the production field. , resulting in less investment in the entire society, less capital reserves, and become poorer in the future.

Statement 2 (iii) What was wasted? fluidity.

Third, we all know that if a certain prover earns enough BTC to buy a Lexus, and the total amount of Lexus in the whole society is quantitative, then one person buys it, and some people can't buy it accordingly. If you want commodity prices to be "cheaper", you must produce more goods (eg, more Lexus).

If there is no mining activity, we can save electricity and save silicon for manufacturing mining chips. Can these saved resources be used in conjunction with other production? of course can. But the benefits of these “non-wasting” resources are offset by losses from other links. PoW directly affects energy and hardware, while Tendermint implicitly affects liquidity.

Let's take a little bit of a circle and talk about the loan interest rate and the market. Suppose you are a hungry and hungry hunter, and suddenly you find a hare that looks very delicious. The demand curve starts to work: your body becomes excited, begins to breathe, and the only remaining energy begins to infuse your muscles; your brain accelerates oxygen supply, and the blood delivers glucose quickly because it is running at speed, how is it planned? Grab this hare. Before that, you don't need to waste the only energy left (funds "liquidity"), but (the brain tells the whole body) we know that if we catch this hare today, we can replenish all the energy we need, and even get a rest. Opportunity, practice new hunting skills, pass on generations, etc. What I want to say is that we can imagine a series of causality, such as hunting, gathering energy, mobilizing biological instinct, and stimulating psychological effects.

In PoW, liquidity resources (silicon, metal, concrete, electricity, time, manpower – each can be used for many things, but the use of any element is limited without any combination) is pooled into devices that cannot flow ( Mining equipment and facilities need to build, manage, maintain equipment, and maintain the social relationship between equipment owners and their business partners; integrating these things can achieve a specific purpose, but they do not make sense to split them. Investors are actually betting that “there will be no more efficient combinations than this manufacturing-production approach”. So for Tendermint-PoW, these certifier deposits are locked out and cannot be used for other purposes. Those who choose to deposit the deposit are actually betting on it; the waste here seems to be unobservable (because it involves a known – but the consequences have not happened), but actually still exist.

In the case of PoW, it is like a hunter grabbing a rabbit while lifting the iron (more difficult to catch the hare), in the case of TenderPoW, just like glucose and feeding can not be delivered to the brain in time (and more difficult to catch the hare); no matter what The way that always leads to the purpose (grabbing the hare) is more difficult to achieve. In fact, the power and interest rate of the market will cause the whole society to find a balance between the two methods (that is, the same rate of return) (once any method is too profitable, it will lead to a large influx of funds, which will lead to an increase in market interest rates; Similarly, consumer projects that are more prone to satisfaction can also cause this consumption).

In the long run, there may be differences between the two PoWs: PoW is good for economic growth, and SoftAltoid is good for reducing carbon dioxide emissions. This is not conclusive, but the reduction in funding means that progress in areas such as technology and technology, storage and emission reduction, and carbon substitution is hindered, which will of course have a negative impact on the future.

Combined with mainstream knowledge

Open your economics notebook, here is the supply and demand curve for marketable funds:


The current situation is that the supply line of loanable funds moves to the left, indicating that the loanable funds are invested in “non-production behavior”.


Areas C and D are the “waste” generated by the verification behavior (deposit).



The benefits originally belonging to Regions C and D were wasted…

Example 2: DPoS DPoS is a kind of chaebol governance system (not democratic, nor capitalist). People use the money to vote for 100 representative representatives. These people sign the block in a certain order. Ensure the security of a network that is still a peer-to-peer.

As always, we need to ignore all of the following issues; the chaebol governance system refers to the same wit (or stupidity) of every decision, and the inevitable fall (unlike "elite" capitalism); DPoS strongly concentrates wealth (controlled by agreement) Voters (even very efficient corporate legal entities) will not pay attention to or understand the motions they support, and any 67 representatives will be able to launch a double-flower attack that cannot be discovered.

But now, I just have to prove that the bitcoin issuance method DPoS and PoW results are still the same.


In theory, the assumption that DPoS can operate safely is: [1] "Every "non-attacker" can know which 100 delegates are suitable for this job…"; [2] "… … to some extent, the system will not be artificially destroyed…"; [3] "…the attacker will never hold more than 50% of the money supply. …..".

[Conclusion] "…this network can guarantee security forever."

It can be seen that [2] is the key to the success of the DPoS system, [1] without any real help.


If you have already guessed what I want to say now, please skip this section and just look at the "Coin-Modify Paradox" section. However, because this happens to be my personal interest, I am prepared to spend more pen and ink to prove that "voting is only effective when the cost of knowing is low" (obviously DPoS does not meet this condition).

Currency and politics

The ballot is also to be bought. And often it is not directly purchased (this will break the voters' valuable self-deception, become the focus of the thousands of people), but through tax relief, welfare, military operations, subsidies, research bonuses, permits and so on. On the table, everyone said that all of this is to serve the citizens (the fact is only a part of the usefulness to the citizens), and the subtext is: "Give the candidate X vote! Equal to the sky off the pie!" In addition, there are ads Promotional spending and management of individual campaigns. Because the election campaign is a zero-sum game, these activities cancel each other and eventually create a “net waste” for the entire society.

Is such a ballot trading a bad thing? Who knows (think of the long-term cooperation problems facing humanity)? I don't care about the purpose of writing this article.

I am not trying to argue that this is a good or a bad thing. I am just proving that my above remarks are true.

The more total votes, the easier it is to buy votes. Although "the more voters, the smaller the weight of each voter's votes" sounds like this, but it implies a profound paradox. In the end, the huge size of the voter will make every voter insignificant – just like a balloon blowing to a certain extent, the plastic on the surface of the balloon will become extremely thin. If only 3 people vote, one vote can break the tie, so any one is the key; but if it is 10 million votes, the draw may not appear at all (so any voter is insignificant) . Just find a small group of “key players in ignorant voters” and instill them with some desperation that leads to self-realization, and you can completely control the entire voting process. If the cost of knowing is low, such self-fulfilling expectations are difficult to manufacture. However, unfortunately, participation in voting is not cheap and requires voters to spend energy, and most voters will not spend this thought. This detail means that politicians who use public resources to support their bribery will always have an overwhelming advantage.

Wealth distribution is not average enough

However, DPoS is not one person, one vote, is to vote with wealth, will it be better?

This kind of improvement is quite limited: although wealth is never evenly distributed (perhaps lognormal distribution or distribution according to the law of power), a project targeting the “global book” may have very A large number of users; the number of users will still be large enough to allow a well-informed vote to be impossible, because each individual will feel that this is a waste of time.

The indifference of voters has undoubtedly hit a more ideal scenario of modern corporate governance. In the shareholding system, the voting rule is one share, one vote, and the interested parties can sell the stock and give up voting (this is not possible in DPoS). The authority to elect the board of directors by shareholders is almost meaningless; most shareholders are not concerned about the company going to vote, nor are they concerned about understanding the obvious corruption or even taking action, like executive compensation issues, poison pill plans, power Expansion, gold parachute and so on are still popular.

DPoS, like all votes, is not capitalist . Capitalism is not "one dollar, one vote", but "take one dollar to take risks, you get a vote." Capitalism requires a one-to-one correspondence between income claims and risk exposure. In voting, the benefits of “quality ballots” are public, equal and widely distributed (and therefore no right to private gains), and the votes themselves are free (and therefore private property is not used to take risks).

Back to workload: scarce cognitive resources and advertising

If you learn to vote for someone who needs to pay “workload”, then even if there is no bribe, voters will always be affected by this “workload”. In the 2012 US federal government election, SuperPACs raised and spent more than $500 million. If not, how can people be willing to pay for it?

Many studies have proven the effectiveness of political advertising (especially "personal attack advertising"). You can go to this website and click on the "users also read" sidebar. Recall that as long as there is a positive correlation between the two, it is enough, no matter how small. It's as if you only get a hash value and it's almost impossible to get a Bitcoin block reward (but you'll keep counting), and only playing an ad is unlikely to win the election. However, the participating parties will decide the total expenditure for the election based on their total return expectations.


Cancel the rent, the rent will not disappear

Trying to create a factor between the cost and the outcome to prevent the two from corresponding is destined to fail; once the benefits are greater than this, the measures are completely broken. (Again, again, you only have a hash in PoW that is unlikely to get you a block reward). For example, in the United States, citizens who do not understand the MC=MR principle may try to limit the amount of political propaganda donations to a certain size. However, because advertising is always effective (MR), MC will pick up sooner or later. Only at this time, the form of MC will become the advertisement of SuperPAC, and the MR is still purchased. However, there is an inconvenient factor that voters cannot legally coordinate with their candidates.

Sometimes these candidates may secretly admire these advertisements, and sometimes they will sigh how there are such pig teammates. We can't verify, we only know that Gingrich lost. If anti-corruption is as simple as “prohibiting political spending”, our society will not have a history of these problems. The fathers of the founding fathers of the United States know that the only reliable anti-corruption measure is to establish a government with limited power. Eliminating MR, MC (expenditure spent on corruption, lobbying, and publicity) will really come down.

I sincerely hope that we can do this, but the next section will prove that we can't do it.

Is voting really feasible?

If the cost of creating a full informed vote is zero, and the number of votes can be properly aggregated (such as averaging in a yes/no binary vote, taking a median value in a numeric vote), voting is possible (even working) Very good).

If voting is simple and easy (for example, with a smartphone or type of stuff) and only for preferences/values/priorities (all things in your own mind, no time-consuming and laborious), then voting can produce some reliability. The signal shows which measure best respects the feelings of most people. This signal is a very powerful, non-violent social organization tool.

However, in our case (verification of the consensus of the blockchain, and the attacker may be longer than obscuring the voters), the so-called fully informed vote will always be expensive.

What is the meaning of the "workload" requirement? I repeat, if the money distribution process does not produce endless zero-sum competition, the distribution plan will completely ignore the participants' efforts.

To achieve this, there must be a super-fascism that can never be subverted. It is only distributed to those who are 18 years old and have X SGD per person. They are forced to move and never see goodbye to friends they knew before the age of 18. . Moreover, even if it is extremely extreme, this system may also give some benefits to those who (selfishly) have many children (they will never see their children again); and also assume that the inflation tax is incompetent. Escaped (ie, every inflation is a perfect one-off tax), assuming that the currency issuance plan/the government itself will never be changed or corrupted.

Therefore, (since it cannot be achieved), in terms of digital currency, we return to the ultimate consideration: the issuance plan.

Mint-corrosion paradox ("less is more")

Reducing the amount of money issued in each block will actually increase the total economic value of each block. Pre-drilling will only motivate people to use multiple chains to compete for benefits.

The real unit of money property is the percentage. Regarding money, what you really care about is "how much money other people have."

If someone can change a Steinway grand piano to me, asking everyone to get 1000 Steinway, I will accept the deal. I will accept the mansion, television, health and magic power regardless of the situation. I may feel disappointed, or I may doubt that I am dreaming, but this time I only need to care about one thing, that is, myself, so I will accept any free things.

However, money is not the case: if the currency is changed out and distributed to others, then my purchasing power will drop. It’s like my money is rusting.

Note that multiplying the number of currencies by a multiple does not change anything (in fact, this is equivalent to changing the unit of the dollar from "dollar" to "cent" and then multiplying all price figures by 100). Any change in the amount of quantity is simply a change from "dollar" to another, not anything.

It is precisely because the uniform quantity change is meaningless, and the uneven quantity of money changes is significant. If you work hard all day, you can get $100. Others print $100 without any effort. You definitely want to ask for a higher price for your job. What you care about is that you get "a few percent of the total amount of money", you want this percentage to be higher, and others want your weight to be lower.

“If we only issue 5 bitcoins (not 50) every 10 minutes? Isn't that thing 10 times cheaper?” No, I don’t think so. Reducing all numbers by a factor of 10 does not change anything (it's like taking dimes instead of dollars to price items).

However, manipulation of price levels is possible by arranging rules for the issuance of new amounts of money.

It’s like you can get $100 a year later, people will prefer to get $100 tomorrow; a dollar holder will feel that printing $100 tomorrow will be more than $100 a year later. pain.

This is the paradox: the “new currency as a percentage of the total currency” issued today, if it gets lower, the monoblock MR seems to have fallen (because the numbers have become smaller); however, by improving in the future The percentage of new currency issued, the rate of currency rust is slowed down, so this currency will become more valuable (higher quality) and will increase MR.

“Like NXT/BitShares/Ripple, we can send all the money in advance. If there is no block reward, is there no MR?” It is not right. This idea is very problematic and it is impossible to change the principle of MC=MR.

I have written before that such a design would have a fatal effect on the network effect of money (see the lower half of my original text). This kind of scheme can't find a counterpart in nature (so it is ridiculous and dangerous). Third, this scheme will focus the currency on a small group of people, making the start-up very turbulent (as the price of the currency rises on the exchange, all early supporters will become super rich and try to sell once) . If someone expects NXT to succeed, then these early supporters have to protect their identity.

However, did we finally achieve MR=MC=0? No "waste"?

Wrong, because the 100% release itself constitutes a huge MR when the project starts, and like other MRs, the MR also causes competition. In other words, the MC=MR problem is compressed into a small creation block and then distributed laterally across multiple blockchains . Because our eyes are only staring at a chain, these workloads are ignored (such as marketing activities that allow a community to switch chains), as if we were referring to the example of equity grinding.

You ask yourself: Are Ripple and Stellar really two different blockchains? I think this double- chain phenomenon is actually a Nothing-at-Stake attack on a high-margin block (the block is a model chain that can be easily forked into a Stllar or Ripple). These competing chains, NXT, BitShares, NEM, etc., really want to "add" "new" "features" to Bitcoin? I think they are desperate, unconstrained and desperate attempts to kick bitcoin out of the so-called "super bitcoinization" and replace it with something else.

Now that there are too many people around us saying "support blockchain, against bitcoin", there are still many people who worry that "it is too late to get on the bus." These concerns have led new users to consider creating their own (new) systems. BitShares ignored NXT's 100% pre-drilling precedent, and Ethereum ignored BitShares. NEM ignores all of these items and may be ignored by latecomers.

The 100% pre-excavation program expelled all those who wanted to develop new things without leaving them with the benefits of developing new features that would be compatible with them in existing systems. 100% pre-drilling kicks out all the innovators and forces them to bite each other for the winner-take-all mode (…if you really want to bet such a black swan, you might as well take the Internet to South Australia) .

This phenomenon of competing for the leading throne in the same blockchain is also a form of rights grinding, but less obvious, is a bureaucratic/marketing-based security.

"Whereas quickly-inflating coins "rot" (lose value) individually, slowly-inflating coins "rot" collectively, as the entire network "walls itself off" before it can mature (and it instead must "work" against its many competitors).

Delaying currency rust The security of blockchains is not the primary function of Bitcoin PoW. Instead, the role of PoW is to delay the issuance of money, so that when potential network participants first discover the project, it is still very cheap (which is why some people complain about the “dumping” problem of miners).

Everyone wants them to invest in bitcoin early on, but the pattern of early bird eating means that the network can't grow. ("Do everyone want to use this network?" "How can the agreement pay for its own marketing fees, development fees, etc?") These questions are far more important than some boring technical issues (such as Byzantine generals) . A broken system is useless, which is obvious, but it is a real problem compared to getting the system up and making the system useful.

in conclusion

The “work proof” exists because money is gradually created. Therefore, it is impossible to “create a new form of currency” without introducing proof of workload.

Nakamoto's design genius is to introduce the inevitable work into a process that can be accumulated, perfectly stabilizing an anti-cartel point-to-point clock.


The concepts in this article have been outlined by many thinkers a long time ago. To write about PoS, it's impossible not to mention Andrew Poelstra and Greg Maxwell; they also laid the groundwork for this broad rebuttal.


Original link:

Author: Paul Sztorc

Translation & Proofreading: IAN LIU, Ajian & Ajian