As cool as blockchain is, it sure is starting to pick up hype. When these.. I don't know what to call them.. culture theorists..? start talking about how it will CHANGE EVERYTHING, I really start to get skeptical.
Look, I think blockchain can basically be seen as a public, distributed-trust ledger. It's not like it doesn't have any kind of social precedent, the concept of a "ledger" has existed... throughout all of modern civilisation? I think the decentralised and computation-oriented nature of blockchain has some interesting uses, but let's not pretend we didn't previously know what a ledger is.
Pretty much all of the things listed in this page under "Web 3.0" are things we already do. Blockchain might eventually lead to new ways of doing these things, but let's not pretend it will CHANGE EVERYTHING, it will just allow to do these things, maybe, a bit more efficiently and with more trust. Maybe.
Sigh. Sorry, already getting a little tired of the utopian predictions starting to be promoted by techno-evangelists jumping on this bandwagon. Enthusiasm is good, hype gets old quickly.
The innovation behind a blockchain-based system isn't the ledger or the distributed nature of the thing. It's the incentive structure behind proof-of-work that gives Bitcoins their value.
The non-zero value of the bitcoin network is the reason the blockchain exists, and it's the reason it can be used as a decentralized neutral third party for building apps. The blockchain "killer app" is money (Bitcoin).
That is the million BTC question. Bitcoin, on its own, could "change the world". Other blockchain apps could be influential even if the network/reach of Bitcoin stays at its current level. It's unclear which implementation (if any) will be the winner in the long run.
Noone wants to buy Bitcoin to enrich those who came before them, so they'd rather setup their own version and sell it to others as "the next big thing"
Thinking about some blockchain precedents can lead to some valuable insights about what we can build with them. It's effectively a scroll. Societies used scrolls to record actions so they could consult history to make decisions. Power was concentrated in the scribes and their masters who controlled the scrolls. If we now have a digital scroll in the cloud controlled by no one, that could be a powerful tool for coordinating entire societies without concentrating power. Maybe.
How about we stop using blockchain like it's a thing on it's own and not a concept of which there are many implementations?
Blockchains are an idea, just like atoms are an idea, we don't refer to them as "atom" when talking about the concept. The #1 way to know someone doesn't know what they're talking about is when they use "blockchain" like this. If you say "the blockchain" in the context of a specific cryptocurrency, that makes sense, just like if you say "the atom" in the context of your scientific paper referring to a specific atom. But you can't just say "blockchain" referring to the concept.
You can say "I understand biology" without saying "I understand the biology of humans". You can say "I understand blockchain" without saying "I understand the bitcoin blockchain".
Think of it as "I understand the biology of human" vs "I understand the biology of humans". One makes sense, the other seems meaningless, not grammatically correct (and so blatantly it sounds terrible to a native speaker) or at least makes you ask which human.
I view the use of the term "blockchain" as a good litmus test as to the speaker's motivation. The word should really only be used to refer specifically to the database structure that is a blockchain. When someone uses the term blockchain to refer to the bitcoin network, bitcoin tokens, or the bitcoin industry or "movement", I suspect that the speaker does so b/c he is uncomfortable saying "bitcoin". He wants to talk about bitcoin, without appearing to endorse the ideals that gave rise to bitcoin. Perhaps he thinks satoshi's approach is misguided, or he eschews satoshi's ideals in favor of making a living from the existing business/state edifice. Plenty of smart people are in this camp, and to them I would charitably allow that they use abuse the term blockchain out of an honest desire to seem less threatening to the uninitiated.
The current title of “Still Don't Get #Blockchain? an Explanation for Five-Year- Old” is rather incorrect anyway. I wouldn’t expect the average five-year-olds to be helped particularly much by this article. Especially when it slings around terms like “P2P” without explanation.
In my experience, 99% of people who talk about the blockchain innovation have absolutely no idea what they're talking about. They also have no idea (or pretend to have no idea) why Bitcoin was invented in the first place: it was invented to avoid the need to ask a permission to send value. That's why a blockchain and proof of work were needed - so people who don't know each other could exchange value on the network safely and without intermediaries. It was not invented to make AML/KYC easier or to make banks richer.
One thing this article and many so called experts don't "get" about the blockchain is that it's pretty useless without the currency, namely Bitcoin. There has to be an incentive for lots of decentralized parties to verify (by spending energy) transactions on the blockchain and agree on the history - that incentive is mining currency that has value. If you "invent" your own blockchain, not many people (if any) are going to mine it, making it de-facto less secure than Bitcoin.
If you have a number of financial institutions (who already trust each other) that need to process transactions, send documents etc. you don't need a blockchain - you need a database and a messaging system. Blockchain is going to be a less effective and more expensive way to do the same job in this case.
I agree that lots of people who talk about blockchain don't understand how it works.
However, there is plenty of cool things to like about a blockchain without POW. A blockchain is just a database which broadcasts signed transactions over a p2p network.
Forging a blockchain requires a validator to sign two conflicting transactions. When that happens, you have undeniable proof of fraud which you can take to a judge.
Compare this to most databases where a sysadmin changes a few values and deletes the logs, no one would be any wiser.
The whole point of a blockchain is to prevent double spending by untrusted parties. If you have to go to a judge to resolve your transaction then you're kind of missing the point of this entire experiment.
> Compare this to most databases where a sysadmin changes a few values and deletes the logs, no one would be any wiser.
To me that screams "use off-site backups with different trust roots not sysadmined by the same sysadmin". Invoking "blockchain" here is unnecessary, and might unnecessarily confuse folks into adopting tech that doesn't solve their underlying problems.
But the average user has no idea whether a system is "using off-site backups with different trust roots not sysadmined by the same sysadmin." And even if it was like that at one time, they have no way of continually auditing the system to confirm that, at all times, that process is in place. And they have no idea that two sysadmins aren't in cahoots with each other.
What this misses for me is that once we step beyond the conventional database, there are many, many things besides blockchains.
So I don't think the right question is, "Is this better than a normal database?" I think it's "Is there a way we can solve the particular user needs with something simpler than a blockchain?"
Given that most of the commercial blockchain uses I've heard of involve some sort of trusted party or legal arrangement, I ask that question a lot. So far, I've never gotten a persuasive answer that blockchains are needed.
> A blockchain is just a database which broadcasts signed transactions over a p2p network.
That is certainly the definition some people seem to be using. However, that's not even remotely the innovative part of Bitcoin. If that definition of "blockchain" is acceptable, you might as well call git a blockchain.
What others consider to be the key innovation in Bitcoin is the combination into a self-feedback system of proof-of-work-mining (variable according to network power as determined by looking back at previous blocks), longest-chain-wins and the ever-growing-ledger with standard-rules-for-ledger-validity. The real genius is that this creates an incentive to mine while making double-spend impossible without 51% of the network (which is made difficult because others have an incentive to mine, and so it goes round).
The Wikipedia definition of "blockchain" appears to be a hashed chain of blocks (so basically git) with the addition of a selection-of-winning-branch algorithm. I'm not sure that's really generally accepted as the definition of "blockchain", though. The term seems to be suffering from inflation at the moment as everyone working on anything related to keeping a distributed ledger jumps onto the hype bandwagon.
This article is no better. It reduces the real innovation by referring to "known algorithms". Thus we don't even know what the author considers a definition of a blockchain to be!
I was pretty excited about tamper evident logging from some work I was doing (medical records). I haven't delved into the P2P/consensus part, but I could see how that might be useful too.
Apart from the currency, I haven't thought of another use case for PoW.
A buddy keeps pestering me about blockchain-based voting schemes, which eliminates the secret ballot (meaning: non-starter), but I got curious any way.
I completely agree with your observations about people not getting these things; another thing that can be difficult is that the blockchain mechanism is not magically incorruptible but has some built-in assumptions about incentives (which indeed rely on mining rewards) and hash power or distribution of the ability to perform some other task. Depending on other real-world facts, these assumptions might be hard to realize or maintain, and there's been ongoing research about that.
Surfacing these assumptions and considering how realistic and sustainable they are would also be valuable for people talking about this kind of technology. For example, what would happen if sophisticated people chose to spend $50,000,000 to attack the system in some way? Or what kinds of conspiracies or coalitions among miners are stable, and what will they be able to do?
I don't mean that these kinds of questions are inherently unsolvable or unanswerable, just that it's another area that people who talk about blockchains may have failed to familiarize themselves with. Or in a way it's an extension of what you're talking about: if people don't realize how blockchain mining is based on a monetary incentive, they might not realize that that incentive and that process may be fragile and subject to some kinds of attacks.
Yes! And these linkbait articles are terribly irresponsible. "Blockchain for 5 year olds" might as well be "complex derivative products for 5 year olds". It is insulting both to our intelligence as adults, and to the intelligence of some very bright 5-year-olds I have met over the years. There are some 5-year-olds who likely understand bitcoin far better than the author of the head article. At the same time, promoting any financial product to people who do not understand (and then selling them a false explanation) is unethical at best.
It's important to understand that buying cryptocurrency powerfully shapes your beliefs. All incentive systems have this property, and since blockchains give us the ability to program with scarcity, most blockchain-based applications can shape your beliefs in powerful ways, too. Knowing this, participating in communities who hold the same incentives you do is almost guaranteed to reinforce false beliefs, like the belief that blockchains are useless without Bitcoin. It's a common belief in the Bitcoin community because the whole community is incentivized to believe it and spread that belief.
If you've ever worked in a startup, you'll see the same incentives at work. Everyone holds equity in the company. Everyone is incentivized to believe that equity will be valuable and to convince others—especially new hires—that the equity will be valuable as well. Even when a company is clearly failing, paths to success will be invented, and belief in those paths will be spread.
If you want to think clearly about blockchains, you have to realize that they can make you crazy, no matter how rational of a person you are. It's not a character flaw. It's the human flaw that makes the incentive systems our society is built on actually work.
There is no evidence that the commenter holds or trades bitcoin, and you also didn't provide any reasons for a blockchain being more valuable than private databases in the contexts he talked about.
Name a product that is widely used as a private, shared database and isn't based on blockchains. It was possible to build these things before Bitcoin, but they were uncommon. Now they're common and people call them private blockchains.
This same conversation will happen over and over because there are thousands of people who must believe that Bitcoin is special. It is special, just not in all the ways people claim.
If the commenter doesn't hold Bitcoin, he's free to deny it. But the comment wasn't about him as an individual. It was about how blockchains make people crazy, including me.
> like the belief that blockchains are useless without Bitcoin
It would be more accurate to describe the community's attitude as "blockchains are useless without a native currency to incentivize network security", which is definitely correct in my opinion. Having a native currency is crucial for the only true innovation in Bitcoin/blockchain - obtaining distributed consensus among non-trusting peers. The various "blockchains" out there without a native currency cannot achieve this goal, and therefore are not really blockchains [0] in my opinion.
[0] "blockchain" is really becoming an empty word nowadays and does not really have a clear definition, so each to his own... for me, "blockchain technology" represents the breakthrough that Satoshi came with - proof-of-work-based distributed consensus. For others, it means an append-only hash-linked data structure, but this existed long before Bitcoin and was possible since forever, and so not very interesting to me.
I am by no means an expert (actually this is all pretty new to me so I look forward to feedback), but that diagram seems a bit dated. There's a lot of potential for scripted blockchains like Ethereum to cut out a lot of transactional overhead in a lot of industries.
In a private blockchain setup there really is no need for incentives - the parties participating in the blockchain are covering the expenses.
Financial institutions do "trust" each other to a higher extent than some random actors but they still maintain ledgers that are getting reconciled and in the vast majority of cases the funds aren't dispersed to their intended recipients until the reconciliation flow completes unless the sending counterparty maintains a balance in recipient's system.
Blockchain implementations take this trust to a different level and as a nice side effect make reconciliation/dispute resolution a breeze compared to a standard setup.
If you know this from experience working on actual reconciliation systems, could you elaborate on why this is the case?
Why is bothering with a blockchain so much better than SWIFT for example? That's a system that has third parties logging transactions so it's not like banks can just say "oh no we didn't"
SWIFT is essentially a messaging system, the funds still have to clear elsewhere. Blockchain solutions (ripple being the biggest crypto for interbank funds exchange atm) combine these 2 systems basically.
With so many moving parts involved with proprietary not immediately visible logging mechanisms - the banks play it safe especially when it comes to international transactions spanning multiple jurisdictions etc.
As long as the blockchain-based currency doesn't have a legal tender status, won't banks have to clear elsewhere anyway?
The way I see it, unless the central bank joins the party and declares records on the blockchain as the final word over ownership and that the blockchain currency must be accepted as legal tender, the blockchain records are just that - historical records to keep track of who-owes-who-what, with settlement happening elsewhere and enforced via contracts, lawyers and court.
The point being, blockchain makes sense for bearer assets like Bitcoin, where owning the key is owning the asset and there are no other external processes that can effect ownership (moving around trucks with cash, court disputes, etc).
What banks really need, in my opinion, is simply a cryptographically-signed append-only log where they record debt among each other. No global consensus, shared state or blockchain is needed here - just a private log for each pair of banks that transact with each other, kept locally for the interested parties only, and which can be used in court if either party misbehaves and does not settle the payment.
I never get an answer to this basic question. There seems to be all this money thrown into the (not bitcoin) blockchain, but nobody is ever willing to explain in detail why that is.
What are the moving parts now, and how exactly will a blockchain make them go away, given that any blockchain a bank would use won't directly represent the assets the bank claims to have? How can it ever be different from an authenticated messaging system?
I'd be very interested to read a detailed, technical explanation.
Check out Ripple docs. That's just one angle of looking at the situation (using a non-100%-privately-controlled-system) but it should clarify a few basic things.
If you're curious to see the comprehension of blockchain by the typical blockchain software developer (or the lack thereof) here's a fairly typical example of how the pitch devolves, once examined: https://www.youtube.com/watch?v=BA2_pg_odX4
While I understand the benefits of blockchain tech, doesn't it waste resources by constantly requiring more and more energy to keep the blockchain running?
While I can't know for sure, I'd be willing to bet that the Bitcoin blockchain consumes less energy then the global State apparatus surrounding fiat currencies & gold.
Think of all the employees, armies, vaults, trucks, printing presses, etc involved in securing the world's government-issued currencies. Don't those waste more and more resources to keep the fiat money system running?
And let's not forget the cost of shooting war or currency war.
That's a feature of any financial system beyond trivial complexity.
> armies
That's a feature of any society.
> vaults, trucks, printing presses
That's a problem with physical currency, and is orthogonal to cryptocurrencies. You don't need trucks for a bank transfer anymore than for bitcoin.
> And let's not forget the cost of shooting war or currency war.
That's, again, a "feature" of all societies. Bitcoin is completely orthogonal.
I'm open to being convinced that Bitcoin is (and will be), in general, less of a waste than the upkeep we're paying for the current system. Because I don't believe so right now, and the additional problem is that Bitcoin literally feels like taking lots of useful fuel and burning it on the side of the road as an offering to Security Gods.
This blog post does a great job explaining some of the economics behind mining and why it is not wasteful, in the context of showing proof-of-stake is nonsensical:
Both are highly recommended read, but a quick TL;DR: if you're going to incentivize the security to of your network with $X of rewards, then up to (and probably very near to) $X would be necessarily "wasted" to obtain that reward. And these incentives and security model are not "waste" in any way - they're the entire premise behind blockchain systems!
edit: removed quotes from truthcoin.info, added tl;dr
Explain this for me, because I don't feel Truthcoin's argument holds:
Truthcoin claims that any money system will cause the exact same resource expenditure, no matter what method you use. That resource cost is equal to the value of the coin. According to the argument, Proof-of-Work is merely the most direct and obvious expenditure of the resource cost.
That may be a fine argument against Proof-of-Stake; I'm not conversant enough with PoS to say. But how does that argument hold compared to the cost of traditional banking?
Truthcoin seems to explain away traditional banking by saying it incurs rent, and "rent is incompatible with p2p currency." Do I have that right?
Even if the "rent" charged by a traditional bank was equal to the dollar-cost of PoW, the resource-expenditure will not be nearly as high. In PoW, the participants are burning fuel to reach the cost (expending resources). In traditional banking, the fee of the trade is what creates the cost. The actual question is, whether the traditional banking infrastructure burns through the same amount of resources as PoW does. Does it?
For that matter, while it's a clean economic argument to say that participants will spend up-to-the-value-of BtC to mine it, that's not a correct measurement of the resource-cost of BtC, because many different parties are mining in competition. Therefore, isn't it more accurate to say that the resource-cost expended is up-to-the-value-of BtC for each participant?
> Therefore, isn't it more accurate to say that the resource-cost expended is up-to-the-value-of BtC for each participant?
That's the same. Each participant is getting bitcoins. Otherwise they wouldn't be mining.
But other than that, I agree with you that Truthcoin's argument is incredibly contorted. They're considering the fact that some people can profit from financial services to be as "wasteful" as spending actual fuel.
I'm unsure whether the cost of resource expenditure really tracks that closely with the BtC value. Let's just assume BtC trade value against USD stays stable forever. For the BTC-value to equal the value of mined resources, If a BtC is worth $100 at time of mining, then the entire network would only be allowed to spend $100 trying to mine it.
Let's say there's only 100 miners right now. If they all had equal mining power, then they'd mine a coin 1/100th of the time, and be incentivized to spend only $1 each per coin. Then this premise would hold.
The problem is, each miner can increase its odds by increasing the resources spent. So, everybody is incentivized to drive their expenditure higher than $1. Any time a miner makes a short term gamble, or finds a way to externalize their resource cost, they'll spend more than $1, and win. For that matter, would a miner have enough information to know that $1 is the correct upper bound? Not only does that require a perfect measure of the entire network's computing power, it assumes BtC is value-stable. The fact that BtC keeps gaining value against the dollar means that miners ought to be speculating their gamble against a potential future value.
That seems to assume a perfectly-efficient currency, one whose value is always directly proportional to the work invested.
But there's no evidence that BitCoin actually works that way. I have no guarantee that I would spend the same cost to mine one Big Mac's-worth of BC as I would to buy a Big Mac, because the price fluctuates wildly due to all sorts of other factors such as hoarding, worries about forking, etc etc.
>In my experience, 99% of people who talk about the blockchain innovation have absolutely no idea what they're talking about.
As someone who falls into this group (but doesn't pretend to know), where do you recommend starting to learn about the philosophy and technology behind bitcoin? There seems to be so much misinformation out there, and your comment makes me even more skeptical. Are there any books worth the time?
Most people selling "private blockchains" are just trying to sell a (hopefully) better database system to a customer that doesn't really understand what a blockchain is. Sometimes they are adding a messaging system to the equation, and sometimes they are just trying to sell another CRM system. But it's definitely not a blockchain.
Furthermore all this constant search about "uses cases for blockchains" is ridiculous. You simply do not go looking for use cases for technology - that's a recipe for unsuccess unless your business is to make money on consulting hours - what you do is to simply build technology to solve specific problems. As you said the blockchain was invented to solve a very specific problem: sending value from A to B without the intervention of a financial institution or any other third party.
Banking is all about ledgers and everybody is rolling their own. That's why there are big clearing houses that allow funds to be transferred between entities. It's a form of minimal protocol, but it's still incredibly inefficient and redundant. Most of this stuff is outsourced and few people in management have an idea on the incurred cost as everyone is rolling their own wheel design.
The thing blockchains promise to traditional banks is to agree on a common ledger protocol (one common wheel type, preferably round). This would allow to cut out some middle man and save cost in the transactions, maybe even make them faster. Not that established banks really want this, but they have to keep the fintechs in bay, so it's a topic.
You're talking about and assuming only a PoW blockchain but there are other consensus technologies such as Practical Byzantine Fault Tolerance which are actually older than Satoshi's PoW.
Absolutely, there is a nice in built feedback loop as well.
Bitcoin increases in value which makes it more desirable, which drives miners to mine more, which increases the security of the network ( harder to attack ).
This drives the value of Bitcoin higher, repeat.
Anyone can have a blockchain, having a secure blockchain is not so easy. Miners secure the network and their reward is Bitcoin.
It feels like a buzzword to me in this context and I don't think this exact term has to be used to be useful. I'd agree that it would be helpful to talk about the distinctive goal of trustlessness/letting people who don't trust each other reach consensus/not having to rely on a central arbiter/reaching consensus under conditions of not knowing who the other participants are/etc.
I think ii is imperative to note that adversarial environments are what blockchains are useful for. They're not particularly efficient as a means of storing data, so if you can trust and/or rely on suing your counterparties, you don't want to conduct your business on a blockchain. Yet the blockchain hype seems to be much about banks setting up their own chains etc, which is complete nonsense as that would be little different from a regular database.
I'm wondering, doesn't the blockchain expend vast amounts of electrical energy to do artificial work just to keep the system credible? And as the capitalization grows, energy consumption grows with it? Isn't this a problem for scalability? For context, bitcoin is 9 billion dollars, global fiat money is somewhere between 30 and 70 trillion dollars.
Yes, that could be a problem for proof-of-work blockchains. Private blockchains (with different threat models) can use algorithms like PBFT that don't spend appreciable energy.
the network expends vast amounts of electrical energy because people are competing to earn bitcoin from that network. they will even do it at a loss because bitcoin is 9 billion dollars, and can be between 30 and 70 trillion dollars.
if they turned off their machines it would spend less energy on aggregate, and the system would be just as credible, because I would turn my machines on.
The blockchain has either proof of work or some signatures attached to it. While git requires you to resolve conflicts manually, a blockchain can resolve them automatically.
So, Microsoft docs is bad, only one edit can be made at a time. That's why we need blockchain, so we can be like Google Docs instead... Only in pretty sure Google docs doesn't use blockchain so the analogy makes no sense.
I still don't get it after reading the article. Although to be honest, I don't want a 5 year old explanation. I want a technical breakdown of how blockchain technology works without any mention of bitcoin or currency or mining. I feel like Bitcoins really cloud my understanding of it and narrows it down to a subset of its use cases. Really I want to understand better how it can be used as a distributed P2P database of transactions and how exactly it has security and history built in. Anyone have resources? Everything I find seems to have only fin-tech in mind.
The problem with the term "blockchain" is its lack of a clear definition. A blockchain in its most basic form is nothing more than a data structure: an inversely linked list where each node n (for n > 0) contains a hash of node n-1. Bitcoin, of course, uses and stores this data structure in a certain way, but that was done to solve a very specific problem; it makes little sense for most other applications, and it is not part of what a "blockchain" itself is.
What about privacy? Do the block chain technology offer any solutions for protecting the privacy of the fact that a transaction has been made. There are many situations where A would want to keep it a secret that a transaction was made with B.
Satoshi's paper is very concise and the best introduction to Bitcoin (or the damn "blockchain technology" bullshit used to raise money for ridiculous startups, if you want to call it that), particularly for the technical audience here on HN.
https://bitcoin.org/bitcoin.pdf
I read it was Accenture. In any case, it's ridiculous. Although I do understand that customers (who don't know any better) are asking for it.
btw. the proper way to do editable data is to write a smart contract that will control how the latest value can be altered and even then the blockchain will contain the entire history.
101 comments
[ 5.2 ms ] story [ 171 ms ] threadhttps://tonyarcieri.com/on-the-dangers-of-a-blockchain-monoc...
Look, I think blockchain can basically be seen as a public, distributed-trust ledger. It's not like it doesn't have any kind of social precedent, the concept of a "ledger" has existed... throughout all of modern civilisation? I think the decentralised and computation-oriented nature of blockchain has some interesting uses, but let's not pretend we didn't previously know what a ledger is.
Pretty much all of the things listed in this page under "Web 3.0" are things we already do. Blockchain might eventually lead to new ways of doing these things, but let's not pretend it will CHANGE EVERYTHING, it will just allow to do these things, maybe, a bit more efficiently and with more trust. Maybe.
Sigh. Sorry, already getting a little tired of the utopian predictions starting to be promoted by techno-evangelists jumping on this bandwagon. Enthusiasm is good, hype gets old quickly.
The non-zero value of the bitcoin network is the reason the blockchain exists, and it's the reason it can be used as a decentralized neutral third party for building apps. The blockchain "killer app" is money (Bitcoin).
Noone wants to buy Bitcoin to enrich those who came before them, so they'd rather setup their own version and sell it to others as "the next big thing"
Blockchains are an idea, just like atoms are an idea, we don't refer to them as "atom" when talking about the concept. The #1 way to know someone doesn't know what they're talking about is when they use "blockchain" like this. If you say "the blockchain" in the context of a specific cryptocurrency, that makes sense, just like if you say "the atom" in the context of your scientific paper referring to a specific atom. But you can't just say "blockchain" referring to the concept.
Think of it as "I understand the biology of human" vs "I understand the biology of humans". One makes sense, the other seems meaningless, not grammatically correct (and so blatantly it sounds terrible to a native speaker) or at least makes you ask which human.
One thing this article and many so called experts don't "get" about the blockchain is that it's pretty useless without the currency, namely Bitcoin. There has to be an incentive for lots of decentralized parties to verify (by spending energy) transactions on the blockchain and agree on the history - that incentive is mining currency that has value. If you "invent" your own blockchain, not many people (if any) are going to mine it, making it de-facto less secure than Bitcoin.
If you have a number of financial institutions (who already trust each other) that need to process transactions, send documents etc. you don't need a blockchain - you need a database and a messaging system. Blockchain is going to be a less effective and more expensive way to do the same job in this case.
Here's a small diagram to help you understand whether you need a blockchain: https://pbs.twimg.com/media/Cn2zMbTWYAAQA6i.png
However, there is plenty of cool things to like about a blockchain without POW. A blockchain is just a database which broadcasts signed transactions over a p2p network.
Forging a blockchain requires a validator to sign two conflicting transactions. When that happens, you have undeniable proof of fraud which you can take to a judge.
Compare this to most databases where a sysadmin changes a few values and deletes the logs, no one would be any wiser.
To me that screams "use off-site backups with different trust roots not sysadmined by the same sysadmin". Invoking "blockchain" here is unnecessary, and might unnecessarily confuse folks into adopting tech that doesn't solve their underlying problems.
That's not a good basis for a robust system.
What this misses for me is that once we step beyond the conventional database, there are many, many things besides blockchains.
So I don't think the right question is, "Is this better than a normal database?" I think it's "Is there a way we can solve the particular user needs with something simpler than a blockchain?"
Given that most of the commercial blockchain uses I've heard of involve some sort of trusted party or legal arrangement, I ask that question a lot. So far, I've never gotten a persuasive answer that blockchains are needed.
That is certainly the definition some people seem to be using. However, that's not even remotely the innovative part of Bitcoin. If that definition of "blockchain" is acceptable, you might as well call git a blockchain.
What others consider to be the key innovation in Bitcoin is the combination into a self-feedback system of proof-of-work-mining (variable according to network power as determined by looking back at previous blocks), longest-chain-wins and the ever-growing-ledger with standard-rules-for-ledger-validity. The real genius is that this creates an incentive to mine while making double-spend impossible without 51% of the network (which is made difficult because others have an incentive to mine, and so it goes round).
The Wikipedia definition of "blockchain" appears to be a hashed chain of blocks (so basically git) with the addition of a selection-of-winning-branch algorithm. I'm not sure that's really generally accepted as the definition of "blockchain", though. The term seems to be suffering from inflation at the moment as everyone working on anything related to keeping a distributed ledger jumps onto the hype bandwagon.
This article is no better. It reduces the real innovation by referring to "known algorithms". Thus we don't even know what the author considers a definition of a blockchain to be!
Apart from the currency, I haven't thought of another use case for PoW.
A buddy keeps pestering me about blockchain-based voting schemes, which eliminates the secret ballot (meaning: non-starter), but I got curious any way.
Surfacing these assumptions and considering how realistic and sustainable they are would also be valuable for people talking about this kind of technology. For example, what would happen if sophisticated people chose to spend $50,000,000 to attack the system in some way? Or what kinds of conspiracies or coalitions among miners are stable, and what will they be able to do?
I don't mean that these kinds of questions are inherently unsolvable or unanswerable, just that it's another area that people who talk about blockchains may have failed to familiarize themselves with. Or in a way it's an extension of what you're talking about: if people don't realize how blockchain mining is based on a monetary incentive, they might not realize that that incentive and that process may be fragile and subject to some kinds of attacks.
And thanks for the flowchart.
If you've ever worked in a startup, you'll see the same incentives at work. Everyone holds equity in the company. Everyone is incentivized to believe that equity will be valuable and to convince others—especially new hires—that the equity will be valuable as well. Even when a company is clearly failing, paths to success will be invented, and belief in those paths will be spread.
If you want to think clearly about blockchains, you have to realize that they can make you crazy, no matter how rational of a person you are. It's not a character flaw. It's the human flaw that makes the incentive systems our society is built on actually work.
This same conversation will happen over and over because there are thousands of people who must believe that Bitcoin is special. It is special, just not in all the ways people claim.
If the commenter doesn't hold Bitcoin, he's free to deny it. But the comment wasn't about him as an individual. It was about how blockchains make people crazy, including me.
It would be more accurate to describe the community's attitude as "blockchains are useless without a native currency to incentivize network security", which is definitely correct in my opinion. Having a native currency is crucial for the only true innovation in Bitcoin/blockchain - obtaining distributed consensus among non-trusting peers. The various "blockchains" out there without a native currency cannot achieve this goal, and therefore are not really blockchains [0] in my opinion.
[0] "blockchain" is really becoming an empty word nowadays and does not really have a clear definition, so each to his own... for me, "blockchain technology" represents the breakthrough that Satoshi came with - proof-of-work-based distributed consensus. For others, it means an append-only hash-linked data structure, but this existed long before Bitcoin and was possible since forever, and so not very interesting to me.
(2014) https://bitcoinmagazine.com/articles/ethereum-next-generatio...
Turing completeness is always more efficiently expressed as a multisig
The right to run an oracle which broadcasts the winner of the SuperBowl is well protected by the 1st amendment.
The right to arbitrate a gambling contract based on the SuperBowl is not as clear.
Financial institutions do "trust" each other to a higher extent than some random actors but they still maintain ledgers that are getting reconciled and in the vast majority of cases the funds aren't dispersed to their intended recipients until the reconciliation flow completes unless the sending counterparty maintains a balance in recipient's system.
Blockchain implementations take this trust to a different level and as a nice side effect make reconciliation/dispute resolution a breeze compared to a standard setup.
Why is bothering with a blockchain so much better than SWIFT for example? That's a system that has third parties logging transactions so it's not like banks can just say "oh no we didn't"
With so many moving parts involved with proprietary not immediately visible logging mechanisms - the banks play it safe especially when it comes to international transactions spanning multiple jurisdictions etc.
The way I see it, unless the central bank joins the party and declares records on the blockchain as the final word over ownership and that the blockchain currency must be accepted as legal tender, the blockchain records are just that - historical records to keep track of who-owes-who-what, with settlement happening elsewhere and enforced via contracts, lawyers and court.
The point being, blockchain makes sense for bearer assets like Bitcoin, where owning the key is owning the asset and there are no other external processes that can effect ownership (moving around trucks with cash, court disputes, etc).
What banks really need, in my opinion, is simply a cryptographically-signed append-only log where they record debt among each other. No global consensus, shared state or blockchain is needed here - just a private log for each pair of banks that transact with each other, kept locally for the interested parties only, and which can be used in court if either party misbehaves and does not settle the payment.
What are the moving parts now, and how exactly will a blockchain make them go away, given that any blockchain a bank would use won't directly represent the assets the bank claims to have? How can it ever be different from an authenticated messaging system?
I'd be very interested to read a detailed, technical explanation.
Using a trusted 3rd party entity to keep everyone honest?
And what properties of the genius of the Bitcoin blockchain are you using for a "private blockchain setup" that git didn't already provide?
also, I'm not specifically talking about a "bitcoin blockchain" but I suppose that's beside the point at this level of argumentation :)
Think of all the employees, armies, vaults, trucks, printing presses, etc involved in securing the world's government-issued currencies. Don't those waste more and more resources to keep the fiat money system running?
And let's not forget the cost of shooting war or currency war.
That's a feature of any financial system beyond trivial complexity.
> armies
That's a feature of any society.
> vaults, trucks, printing presses
That's a problem with physical currency, and is orthogonal to cryptocurrencies. You don't need trucks for a bank transfer anymore than for bitcoin.
> And let's not forget the cost of shooting war or currency war.
That's, again, a "feature" of all societies. Bitcoin is completely orthogonal.
I'm open to being convinced that Bitcoin is (and will be), in general, less of a waste than the upkeep we're paying for the current system. Because I don't believe so right now, and the additional problem is that Bitcoin literally feels like taking lots of useful fuel and burning it on the side of the road as an offering to Security Gods.
> That's a feature of any society.
> > And let's not forget the cost of shooting war or currency war.
> That's, again, a "feature" of all societies. Bitcoin is completely orthogonal.
No. War is a feature of the State. Currency war is a feature of State control of money.
Seriously, human societies self-organize, and for a good reason.
No. Not even close, if you're counting per transaction. And you can't sanely claim that Bitcoin would obsolete things like armies and vaults.
The only sense in which Bitcoin consumes less energy than money is that fewer people use Bitcoin.
"Long Live Proof-of-Work, Long Live Mining" http://www.truthcoin.info/blog/pow-and-mining/
And the follow-up post:
"Nothing is Cheaper than Proof of Work" http://www.truthcoin.info/blog/pow-cheapest/
Both are highly recommended read, but a quick TL;DR: if you're going to incentivize the security to of your network with $X of rewards, then up to (and probably very near to) $X would be necessarily "wasted" to obtain that reward. And these incentives and security model are not "waste" in any way - they're the entire premise behind blockchain systems!
edit: removed quotes from truthcoin.info, added tl;dr
Truthcoin claims that any money system will cause the exact same resource expenditure, no matter what method you use. That resource cost is equal to the value of the coin. According to the argument, Proof-of-Work is merely the most direct and obvious expenditure of the resource cost.
That may be a fine argument against Proof-of-Stake; I'm not conversant enough with PoS to say. But how does that argument hold compared to the cost of traditional banking?
Truthcoin seems to explain away traditional banking by saying it incurs rent, and "rent is incompatible with p2p currency." Do I have that right?
Even if the "rent" charged by a traditional bank was equal to the dollar-cost of PoW, the resource-expenditure will not be nearly as high. In PoW, the participants are burning fuel to reach the cost (expending resources). In traditional banking, the fee of the trade is what creates the cost. The actual question is, whether the traditional banking infrastructure burns through the same amount of resources as PoW does. Does it?
For that matter, while it's a clean economic argument to say that participants will spend up-to-the-value-of BtC to mine it, that's not a correct measurement of the resource-cost of BtC, because many different parties are mining in competition. Therefore, isn't it more accurate to say that the resource-cost expended is up-to-the-value-of BtC for each participant?
That's the same. Each participant is getting bitcoins. Otherwise they wouldn't be mining.
But other than that, I agree with you that Truthcoin's argument is incredibly contorted. They're considering the fact that some people can profit from financial services to be as "wasteful" as spending actual fuel.
Let's say there's only 100 miners right now. If they all had equal mining power, then they'd mine a coin 1/100th of the time, and be incentivized to spend only $1 each per coin. Then this premise would hold.
The problem is, each miner can increase its odds by increasing the resources spent. So, everybody is incentivized to drive their expenditure higher than $1. Any time a miner makes a short term gamble, or finds a way to externalize their resource cost, they'll spend more than $1, and win. For that matter, would a miner have enough information to know that $1 is the correct upper bound? Not only does that require a perfect measure of the entire network's computing power, it assumes BtC is value-stable. The fact that BtC keeps gaining value against the dollar means that miners ought to be speculating their gamble against a potential future value.
But there's no evidence that BitCoin actually works that way. I have no guarantee that I would spend the same cost to mine one Big Mac's-worth of BC as I would to buy a Big Mac, because the price fluctuates wildly due to all sorts of other factors such as hoarding, worries about forking, etc etc.
As someone who falls into this group (but doesn't pretend to know), where do you recommend starting to learn about the philosophy and technology behind bitcoin? There seems to be so much misinformation out there, and your comment makes me even more skeptical. Are there any books worth the time?
In terms of philosophy, I grasped Bitcoin only after I read enough of the Austrian school of economics.
Most people selling "private blockchains" are just trying to sell a (hopefully) better database system to a customer that doesn't really understand what a blockchain is. Sometimes they are adding a messaging system to the equation, and sometimes they are just trying to sell another CRM system. But it's definitely not a blockchain.
Furthermore all this constant search about "uses cases for blockchains" is ridiculous. You simply do not go looking for use cases for technology - that's a recipe for unsuccess unless your business is to make money on consulting hours - what you do is to simply build technology to solve specific problems. As you said the blockchain was invented to solve a very specific problem: sending value from A to B without the intervention of a financial institution or any other third party.
The thing blockchains promise to traditional banks is to agree on a common ledger protocol (one common wheel type, preferably round). This would allow to cut out some middle man and save cost in the transactions, maybe even make them faster. Not that established banks really want this, but they have to keep the fintechs in bay, so it's a topic.
Bitcoin increases in value which makes it more desirable, which drives miners to mine more, which increases the security of the network ( harder to attack ).
This drives the value of Bitcoin higher, repeat.
Anyone can have a blockchain, having a secure blockchain is not so easy. Miners secure the network and their reward is Bitcoin.
the network expends vast amounts of electrical energy because people are competing to earn bitcoin from that network. they will even do it at a loss because bitcoin is 9 billion dollars, and can be between 30 and 70 trillion dollars.
if they turned off their machines it would spend less energy on aggregate, and the system would be just as credible, because I would turn my machines on.
Ant farms work like real farms does not mean real farms use ant farms.
https://www.reddit.com/r/explainlikeimfive/comments/j3ick/el...
A lot of articles talk about the blockchain without giving any idea about the different implementations.
btw. the proper way to do editable data is to write a smart contract that will control how the latest value can be altered and even then the blockchain will contain the entire history.