1. 28
  1.  

  2. 6

    I think one feature that bitcoin has over other consensus algorithms is that it’s far simpler than any that came before it. The blockchain based on POW is simple. It’s simplicity is what I admire about it.

    1. 4

      As usual, David apparently fails or refuses to understand how and why PoW is useful and must attack it at every opportunity (using his favorite rhetorical technique of linking negatively connoted phrases to vaguely relevant websites).

      That said, the article reminds me of a fun story - I went to a talk from a blockchain lead at <big bank> a while back and she related that a primary component of her job was assuring executives that, in fact, they did not need a blockchain for <random task>. This had become such a regular occurrence that she had attached this image to her desk.

      1. 10

        What would you consider a useful situation for PoW? In the sense that no other alternative could make up for the advantages in some real life use-case?

        But otherwise, and maybe it’s just me, since I agree wuth his premise, but I see @David_Gerard as taking the opposite role of popular blockchain (over-)advocates, who claim that the technology is the holy grail for far too many problems. Even if one doesn’t agree with his conclusions, I enjoy reading his articles, and find them very informative, since he doesn’t just oppose blockchains from a opinion-based position, but he also seems to have the credentials to do so.

        1. 1

          Relying to @gerikson as well. I personally believe that decentralization and cryptographically anchored trust are extremely important (what David dismissively refers to as “conspiracy theory economics”). We know of two ways to achieve this: proof of work, and proof of stake. Proof of stake is interesting but has some issues and trade-offs. If you don’t believe that PoW mining is some sort of anti-environmental evil (I don’t) it seems to generally offer better properties than PoS (like superior surprise-fork resistance).

          1. 13

            I personally believe that decentralization and cryptographically anchored trust are extremely important

            I personally also prefer decentralised or federalised systems, when they have a practical advantage over a centralized alternative. But I don’t see this to be the case with most application of the blockchain. Bitcoin, as a prime example, to my knowledge is too slow, too inconvenient, too unstable and too resource hungry to have a practical application, as a real substitute for money, either digital or virtual. One doesn’t have the time to wait 20m or more whenever one pays for lunch or buys some chewing gum at a corner shop, just because some other transactions got picked up first by a miner. It’s obviously different when you want to do something like micro-donations or buying illegal stuff, but I just claim that this isn’t the basis of a modern economy.

            Cryptography is a substitute for authority, that is true, but I don’t belive that this is always wanted. Payments can’t be easily reveresed, addresses mean nothing, clients might loose support because the core developers arbitrarily change stuff. (I for example am stuck with 0.49mBTC from an old Electrum client, and I can’t do anything with it, since the whole system is a mess, but that’s rather unrelated.) This isn’t really the dynamic basis which capitalism has managed to survive on for this long. But even disregarding all of this, it simply is true that bitcoin isn’t a proper decentralized network like BitTorrent. Since the role of the wallet and the miner is (understandably) split, these two parts of the network don’t scale equally. In China gigantic mining farms are set up using specialized hardware to mine, mine, mine. I remember reading that there was one farm that predominated over at least 10% of the total mining power. All of this seems to run contrary to the proclaimed ideals. Proof of Work, well “works” in the most abstract sense, that it produces the intended results on one side, at the cost of disregarding everything can be disregarded, irrespective of whether it should be or not. And ultimately I prioritise other things over an anti-authority fetish, as do most people -which reminds us that even if everything I said is false that Bitcoin just doesn’t have the adoption to be significant enough to anyone but Crypto-Hobbiests, Looney Libertarians and some soon-to-fail startups in Silicon Valley.

            1. 5

              there was one farm that predominated over at least 10% of the total mining power

              There was one pool that was at 42% of the total mining power! such decentralization very security

                1. 5

                  To be fair, that was one pool consisting of multiple miners. What I was talking about was a single miner controlling 10% of the total hashing power.

                  1. 7

                    That’s definitely true.

                    On the other hand, if you look at incident reports like https://github.com/bitcoin/bips/blob/master/bip-0050.mediawiki — the pool policies set by the operators (often a single person has this power for a given pool) directly and significantly affect the consensus.

                    Ghash.io itself did have incentives to avoid giving reasons for accusations that would tank Bitcoin, but being close to 50% makes a pool a very attractive attack target: take over their transaction and parent-block choice, and you take over the entire network.

                2. 0

                  But I don’t see this to be the case with most application of the blockchain.

                  Then I would advise researching it.

                  One doesn’t have the time to wait 20m or more whenever one pays for lunch or buys some chewing gum at a corner shop

                  Not trying to be rude, but it’s clear whenever anyone makes this argument that they don’t know at all how our existing financial infrastructure works. In fact, it takes months for a credit card transaction to clear to anything resembling the permanence of a mined bitcoin transaction. Same story with credit cards.

                  Low-risk merchants (digital goods, face-to-face sales, etc.) rarely require the average 10 minute (not sure where you got 20 from) wait for a confirmation.

                  If you do want permanence, Bitcoin is infinitely superior to any popular payment mechanism. Look into the payment limits set by high-value fungible goods dealers (like gold warehouses) for bitcoin vs. credit card or check.

                  Bitcoin just doesn’t have the adoption to be significant enough to anyone but Crypto-Hobbiests, Looney Libertarians and some soon-to-fail startups in Silicon Valley.

                  Very interesting theory - do you think these strawmen you’ve put up have collective hundreds of billions of dollars? As an effort barometer, are you familiar with the CBOE?

                  1. 10

                    Please try to keep a civil tone here.

                    Also, it’s hard to buy a cup of coffee or a steam game or a pizza with bitcoin. Ditto stocks.

                    1. -4

                      It’s hard to be nice when the quality of discourse on this topic is, for some reason, abysimally low compared to most technical topics on this site. It feels like people aren’t putting in any effort at all.

                      For example, why did you respond with this list of complete non-sequiturs? It has nothing to do with what we’ve been discussing in this thread except insofar as it involves bitcoin. I feel like your comments are normally high-effort, so what’s going on? Does this topic sap people’s will to think carefully?

                      (Civility is also reciprocal, and I’ve seen a lot of childish name-calling from the people I’m arguing with in this thread, including the linked article and the GP.)

                      Beyond the fact that this list is not really relevant, it’s also not true; you could have just searched “bitcoin <any of those things>” and seen that you can buy any of those things pretty easily, perhaps with a layer of indirection (just as you need a layer of indirection to buy things in the US if you already have EUR). In that list you gave, perhaps the most interesting example in bitcoin’s disfavor is Steam; Steam stopped accepting bitcoin directly recently, presumably due to low interest. However, it’s still easy to buy games from other sources (like Humble) with BTC.

                      1. 6

                        IMO, your comments are not very inspiring for quality. As someone who does not follow Bitcoin or the Blockchain all that much, I have not felt like any of your comments addressed anyone else’s comments. Instead, I have perceived you as coming off as defensive and with the attitude of “if you don’t get it you haven’t done enough research because I’m right” rather than trying to extol the virtues of the blockchain. Maybe you aren’t interested in correcting any of what you perceive as misinformation on here, and if so that’s even worse.

                        For example, I do not know of any place I can buy pizza with bitcoin. But you say it is possible, but perhaps with a layer of indirection. I have no idea what this layer of indirection is and you have left it vague, which does not lend me to trusting your response.

                        In one comment you are very dismissive of people’s Bitcoins getting hacked, but as a lay person, I see news stories on this all the time with substantial losses and no FDIC, so someone like me considers this a major issue but you gloss over it.

                        Many of the comments I’ve read by you on this thread are a similar level of unhelpful, all the while claiming the person you’re responding to is some combination of lazy or acting dumb. Maybe that is the truth but, again, as an outsider, all I see is the person defending the idea coming off as kind of a jerk. Maybe for someone more educated on the matter you are spot on.

                        1. 5

                          There is a religious quality to belief in the blockchain, particularly Bitcoin. It needs to be perfect in order to meet expectations for it: it can’t be “just” a distributed database, it has to be better than that. Bitcoin can’t be “just” a payment system, it has to be “the future of currency.” Check out David’s book if you’re interested in more detail.

                    2. 8

                      In fact, it takes months for a credit card transaction to clear to anything resembling the permanence of a mined bitcoin transaction. Same story with credit cards.

                      But I don’t have to wait months for both parties to be content the transaction is successful, only seconds, so this is really irrelevant to the point you are responding to, which is that if a Bitcoin transaction takes 10m to process then I heave to wait 10m for my transaction to be done, which people might not want to do.

                      1. -1

                        Again, as I said directly below the text you quoted, most merchants don’t require you to wait 10 minutes - only seconds.

                      2. 5

                        Then I would advise researching it.

                        It is exactly because I looked into the inner workings of Bitcoin and the Blockchain - as a proponent I have to mention - that I became more and more skeptical about it. And I still do support various decentralized and federated systems: BitTorrent, IPFS, (proper) HTTP, Email, … but just because the structure offers the possibility for a decentralized network, doesn’t have to mean that this potential is realized or that it is necessarily superior.

                        Not trying to be rude, but it’s clear whenever anyone makes this argument that they don’t know at all how our existing financial infrastructure works. In fact, it takes months for a credit card transaction to clear to anything resembling the permanence of a mined bitcoin transaction. Same story with credit cards.

                        The crucial difference being that, let’s say the cashier nearly instantaneously hears a some beep and knows that it isn’t his responsibility, nor that of the shop, to make sure that the money is transfered. The Bank, the credit card company or whoever has signed a binding contract lining this technical part of the process out to be what they have to care about, and if they don’t, they can be sued since there is an absolute regulatory instance - the state - in the background. This mutual delegation of trust, gives everyone a sense of security (regardless of how true or false it is) that makes people spend money instead of hording it, investing into projects instead of trading it for more secure assets. Add Bitcoins aforementioned volatileness, and no reasonable person would want to use it as their primary financial medium.

                        If you do want permanence, Bitcoin is infinitely superior to any popular payment mechanism.

                        I wouldn’t conciser 3.3 to 7 transactions per second infinitely superior to, for example Visa with an average of 1,700 t/s. Even it you think about it, there are far more that just 7 purchases being made a second around the whole world for this to be realistically feasible. But on the other side, as @friendlysock Bitcoin makes up for it by not having too many things you can actually buy with it: The region I live in has approximately a million or something inhabitants, but according to CoinMap even by the most generous measures, only 5 shops (withing a 30km radius) accepting it as a payment method. And most of those just offer it to promote themselves anyway.

                        Very interesting theory - do you think these strawmen you’ve put up have collective hundreds of billions of dollars? As an effort barometer, are you familiar with the CBOE?

                        (I prefer to think about my phrasing as a exaggeration and a handful of other literary deviced, instead of a fallacy, but never mind that) I’ll give you this: It has been a while since I’ve properly engaged with Bitcoin, and I was always more interested in the technological than the economical side, since I have a bit of an aversion towards libertarian politics. And it might be true that money is invested, but that still doesn’t change anything about all the other issues. It remains a bubble, a volatile, unstable, unpredictable bubble, and as it is typical for bubbles, people invest disproportional sums into it - which in the end makes it a bubble.

                        1. 0

                          The crucial difference being that, let’s say the cashier nearly instantaneously hears a some beep and knows that it isn’t his responsibility, nor that of the shop, to make sure that the money is transfered.

                          Not quite. The shop doesn’t actually have the money. The customer can revoke that payment at any time in the next 90 or 180 days, depending. Credit card fraud (including fraudulent chargebacks) is a huge problem for businesses, especially online businesses. There are lots of good technical articles online about combatting this with machine learning which should give you an idea of the scope of the problem.

                          makes people spend money instead of hording it,

                          Basically any argument of this form (including arguments for inflation) don’t really make sense with the existence of arbitrage.

                          Add Bitcoins aforementioned volatileness, and no reasonable person would want to use it as their primary financial medium.

                          So it sounds like it would make people… spend money instead of hoarding it, which you were just arguing for?

                          I wouldn’t conciser 3.3 to 7 transactions per second infinitely superior to, for example Visa with an average of 1,700 t/s.

                          https://lightning.network

                          as @friendlysock Bitcoin makes up for it by not having too many things you can actually buy with it

                          This is just patently wrong. The number of web stores that take Bitcoin directly is substantial (both in number and traffic volume), and even the number of physical stores (at least in the US) is impressive given that it’s going up against a national currency. How many stores in the US take even EUR directly?

                          Anything you can’t buy directly you can buy with some small indirection, like a BTC-USD forex card.

                          It remains a bubble, a volatile, unstable, unpredictable bubble

                          It’s certainly volatile, and it’s certainly unstable, but it may or may not be a bubble depending on your model for what Bitcoin’s role in global finance is going to become.

                          1. 5

                            Not quite. The shop doesn’t actually have the money. The customer can revoke that payment at any time in the next 90 or 180 days, depending

                            You’ve still missed my point - it isn’t important if the money has been actually transfered, but that there is trust that a framework behind all of this will guarantee that the money will be there when it has to be, as well as a protocol specifying what has to be done if the payment is to be revoked, if a purchase wishes to be undone, etc.

                            Credit card fraud (including fraudulent chargebacks) is a huge problem for businesses, especially online businesses.

                            Part of the reason, I would suspect is that the Internet was never made to be a platform for online businesses - but I’m not going to deny the problem, I’m certainly not a defender of banks and credit card companies - just an opponent of Bitcoin.

                            Basically any argument of this form (including arguments for inflation) don’t really make sense with the existence of arbitrage.

                            Could you elaborate? You have missed my point a few times already, so I’d rather we understand each other instead of having two monologues.

                            So it sounds like it would make people… spend money instead of hoarding it, which you were just arguing for?

                            No, if it’s volatile people either won’t buy into it in the first place. And if a currency is unstable, like Bitcoin inflating and deflating all the time, people don’t even know what do do with it, if it were their main asset (which I was I understand you are promoting, but nobody does). I doubt it will ever happen, since if prices were insecure, the whole economy would suffer, because all the “usual” incentives would be distorted.

                            https://lightning.network

                            I haven’t heard of this until you mentioned it, but it seems like it’s quite new, so time has to test this yet-another-bitcoin-related project that has popped up. Even disregarding that it will again need to first to make a name of it self, then be accepted, then adopted, etc. from what I gather, it’s not the ultimate solution (but, I might be wrong), especially since it seems to encourage centralization, which I believe is what you are so afraid of.

                            This is just patently wrong. The number of web stores that take Bitcoin directly is substantial (both in number and traffic volume),

                            Sure, there might be a great quantity of shops (as I mentioned, who use Bitcoin as a medium to promote themselves), but I, and from what I know most people, don’t really care about these small, frankly often dodgy online shops. Can I use it to pay directly on Amazon? Ebay? Sure, you can convert it back and forth, but all that means it that Bitcoin and other crypto currencies are just an extra step for life stylists and hipster, with no added benefit. And these shops don’t even accept Bitcoin directly, to my knowledge always just so they can convert it into their national currency - i.e. the one they actually use and Bitcoins value is always compared to. What is even Bitcoin without the USD, the currency it hates but can’t stop comparing itself to?

                            and even the number of physical stores (at least in the US) is impressive given that it’s going up against a national currency.

                            The same problems apply as I’ve already mentioned, but I wonder: have you actually ever used Bitcoin to pay in a shop? I’ve done it once and it was a hassle - in the end I just bought it with regular money like a normal person because it was frankly too embarrassing to have the cashier have to find the right QR code, me to take out my phone, wait for me got get an internet connection, try and scan the code, wait, wait, wait…. And that is of course only if you want to make the trip to buy for the sake of spending money, and decide to make a trip to some place you’d usually never go to buy something you don’t even need.

                            Ok when you’re buying drugs online or doing something with microdonations, but otherwise… meh.

                            How many stores in the US take even EUR directly?

                            Why should they? And even if they do, they convert it back to US dollars, because that’s the common currency - there isn’t really a point in a currency (one could even question if it is one), when nobody you economically interact with uses it.

                            Anything you can’t buy directly you can buy with some small indirection, like a BTC-USD forex card.

                            So a round-about payment over a centralized instance - this is the future? Seriously, this dishonesty of Bitcoin advocates (and Libertarians in general) is why you guys are so unpopular. I am deeply disgusted that I have ever advocated for this mess.

                            It’s certainly volatile, and it’s certainly unstable, but it may or may not be a bubble depending on your model for what Bitcoin’s role in global finance is going to become.

                            So you admit that is has none of the necessary preconditions to be a currency… but for some reason it will… do what exactly? If you respond to anything I mentioned here, at least tell me this: What is your “model” for what Bitcoin’s role is going to be?

                    3. 14

                      Why don’t you believe it is anti-enviromental? It certainly seems to be pretty power hungry. In fact it’s hunger for power is part of why it’s effective. All of the same arguments about using less power should apply.

                      1. -1

                        Trying to reduce energy consumption is counterproductive. Energy abundance is one of the primary driving forces of civilizational advancement. Much better is to generate more, cleaner energy. Expending a few terrawatts on substantially improved economic infrastructure is a perfectly reasonable trade-off.

                        Blaming bitcoin for consuming energy is like blaming almond farmers for using water. If their use of a resource is a problem, you should either get more of it or fix your economic system so externalities are priced in. Rationing is not an effective solution.

                        1. 10

                          on substantially improved economic infrastructure

                          This claim definitely needs substantiation, given that in practice bitcoin does everything worse than the alternatives.

                          1. -1

                            bitcoin does everything worse than the alternatives.

                            Come on David, we’ve been over this before and discovered that you just have a crazy definition of “better” explicitly selected to rule out cryptocurrencies.

                            Here’s a way Bitcoin is better than any of its traditional digital alternatives; bitcoin transactions can’t be busted. As you’ve stated before, you think going back on transactions at the whim of network operators is a good thing, and as I stated before I think that’s silly. This is getting tiring.

                            A few more, for which you no doubt have some other excuse for why this is actually a bad thing; Bitcoin can’t be taken without the user’s permission (let me guess; “but people get hacked sometimes”, right?). Bitcoin doesn’t impose an inflationary loss on its users (“but what will the fed do?!”). Bitcoin isn’t vulnerable to economic censorship (don’t know if we’ve argued about this one; I’m guessing you’re going to claim that capital controls are critical for national security or something.). The list goes on, but I’m pretty sure we’ve gone over most of it before.

                            I’ll admit that bitcoin isn’t a panacea, but “it does everything worse” is clearly a silly nonsensical claim.

                          2. 4

                            Reducing total energy consumption may or may not be counterproductive. But almost every industry I can name has a vested interest in being more power efficient for it’s particular usage of energy. The purpose of a car isn’t to burn gasoline it is to get people places. If it can do that with less gasoline people are generally happier with it.

                            PoW however tries to maximizes power consumption, via second order effects , with the goal of making it expensive to try to subvert the chain. It’s clever because it leverages economics to keep it in everyone’s best interest to not fork but it’s not the same as something like a car where reducing energy consumption is part of the value add.

                            I think that this makes PoW significantly different than just about any other use of energy that I can think of.

                            1. 3

                              Indeed. The underlying idea of Bitcoin is to simulate the mining of gold (or any other finite, valuable resource). By ensuring that an asset is always difficult to procure (a block reward every 10 minutes, block reward halving every 4 years), there’s a guard against some entity devaluing the currency (literally by fiat).

                              This means of course that no matter how fast or efficient the hardware used to process transactions becomes, the difficulty will always rise to compensate for it. The energy per hash calculation has fallen precipitously, but the number of hash calculations required to find a block has risen to compensate.

                        2. 6

                          We’ve been doing each a long time without proof of work. There’s lots of systems that are decentralized with parties that have to look out for each other a bit. The banking system is an example. They have protocols and lawyers to take care of most problems. Things work fine most of the time. There are also cryptographically-anchored trust systems like trusted timestamping and CA’s who do what they’re set up to do within their incentives. If we can do both in isolation without PoW, we can probably do both together without PoW using some combination of what’s already worked.

                          I also think we haven’t even begun to explore the possibilities of building more trustworthy charters, organizational incentives, contracts, and so on. The failings people speak of with centralized organizations are almost always about for-profit companies or strong-arming governments whose structure, incentives, and culture is prone to causing problems like that. So, maybe we eliminate root cause instead of tools root cause uses to bring problems since they’ll probably just bring new forms of problems. Regulations, disruption, or bans of decentralized payment is what I predicted would be response with some reactions already happening. They just got quite lucky that big banks like Bank of America got interested in subverting it through the legal and financial system for their own gains. Those heavyweights are probably all that held the government dogs back. Ironically, the same ones that killed Wikileaks by cutting off its payments.

                      2. 8

                        In what context do you view proof-of-work as useful?

                        1. 11

                          You have addressed 0 of the actual content of the article.

                        2. -3

                          Proof of Work can be useful as follows. Blockchain systems are a realization of the P=NP problem in theoretical computer science. A problem is in P if it is intuitively easy to solve on a single conventional machine. A problem is in NP if to check an answer to the problem is intuitively easy to do on a single machine.

                          Now many problems are in NP, cracking hashes is only one of them. If you have a huge network of cooperating computers, then solving a difficult problem also solves consensus. Compare this with a group of schoolchildren; those who have the answers to the test, everybody likes and listens to. But these answers are very hard to attain.

                          Now if P=NP, then we can observe this empirically from the networks’ converging behavior. If otherwise, we won’t observe a convergence, we can’t reject the hypothesis that P!=NP (which is what most scientists believe).

                          Blockchain is a massive theoretical experiment, and to pay for it, we have popularized it and attained a critical mass of followers.

                          1. 5

                            Now if P=NP, then we can observe this empirically from the networks’ converging behavior. If otherwise, we won’t observe a convergence, we can’t reject the hypothesis that P!=NP (which is what most scientists believe).

                            what do you mean by “convergence”?

                            1. -2

                              So, as the popularity of a blockchain increases, the number of participants also increase. To keep solving the consensus problem, the difficulty is increased too. If, however, the difficulty of the PoW improves without an increase of participants, another phenomenon must be at play.

                              If the network at some point becomes unstable (viz. unfairness of consensus) because a few nodes have apparently a superior solution generator, we observe a bias towards particular participants. If, however, such solution generator is hard or impossible to attain, we wont observe this bias.

                              Convergence is that, ultimately, there are only two stable states: stability under participants/difficulty ratio, or, instability from the incommensurability of participants and difficulty.

                              Of course, there are many subtleties in this argument, requiring careful analysis. Consider (false positive): one participant actually owns a mining farm with ASICs, giving it a disproportionate amount of computing power, is that a reasonable explanation of bias? How about scenario’s such as the one of Enigma’s code breaking (false negative), where participants with a superior solution generator won’t tell the others by keeping their advantage to within bounds of reasonable chance, to prevent other parties from inferring the information that the actual problem is already solved. Then can we still observe a bias?

                              1. 2

                                The hashing problem that Bitcoin is based on is not known to be NP-Complete, so even if a polynomial-time algorithm for that were discovered, it would amount to a security hole in SHA256, not a proof that P = NP.

                                On the other hand if someone did have a constructive proof that P = NP, they would be able to use it to break SHA256 (if it were a small enough polynomial with small enough constants). So the Bitcoin network does provide another mechanism of reward for someone who comes up with a constructive proof for P = NP. But the stability of Bitcoin doesn’t add much evidence that P =/= NP, as IMO the other rewards for proving P = NP are more than enough to motivate researchers.

                                Granted, the Bitcoin network does provide further security testing for the crytographic primitives it’s based on, as any high-stakes cryptosystem does.