Lightning Network: “Bitcoin Doesn’t Scale” [pdf]
lightning.network
lightning.network
First there was Gnutella (as Napster had a centralized directory) - but it didn't scale, although it proved popular.
Then they added support to Gnutella for spoke/hub models, which had superpeers that reduced the load on most individuals. But in the end Gnutella never really took off.
What replaced Gnutella was bittorrent, which actually only did transfer via P2P but the searching and discovery was centralized again PirateBay and Trackers.
But for all bittorrent's success, Netflix (and others), which are centralized and highly controlled alternatives, are far more popular and commercially successful for watching online video -- it is just less work to use Netflix, even though it doesn't have the benefits of being free and open.
So for all the attraction of P2P decentralization, often centralized, commercial focused easy solutions win out. There are costs to be paid by being decentralized and often general people who are not ideological do not want to pay these costs.
But I am mostly an outsider with regards to Bitcoin so who knows.
The inefficiencies in the Bitcoin protocol are pretty big by comparison; every full node has to maintain a copy of the entire blockchain, the proof of work scheme is terribly redundant and wasteful. Some partial solutions (peercoin, ripple/stellar) have arisen in the crypto space but they just don't have the popularity of Bitcoin, it'll be interesting to see where the field stands in even five years. The real world demand for decentralized digital currency is still questionable in my mind.
This is a key point. I believe that general people do not care about decentralized currency, what they want are easy payment solutions. If bitcoin doesn't provide easy payment solutions that are cheap, convenient, works everywhere, then it will be beat by something that does provide that.
Even if you could with huge engineering effort slowly move Bitcoin to cheap, easy, everywhere, it is likely going to be beat by non-decentralized solutions because those non-decentralized solutions are cheaper to create, deploy and maintain -- also easier to monetize.
I think that the future is to companies like Stripe, ApplePay -- technologies which focus on customer capture via ease of use and many options and being agnostic towards the underlying technologies.
Bitcoin is neat, just like Gnutella was, but being neat doesn't correlate with commercial success.
And if it is illegal just because the actual money is somewhat anonymous (which Bitcoin isn't, really) doesn't mean that the change in material wealth is unnoticeable.
"BTC" is just the unit of account powering the blockchain. You can use a blockchain to make almost any form of digitized information into a public good. Case in point: https://www.blockscan.com/asset
The blockchain is the ultimate distributed database. No other distributed database works without a third party controlling and administering it. This is critically important for all kinds of financial markets, because otherwise the entity in control of the DB is able to manipulate outcomes, hide data, and play games with insider information, and yes this is really happening: http://deepcapture.com
And why does a purely online-only, digital content provider need your home address to send you a video file, anyway? What about a server host? Gaming platform? The credit card paradigm is obsolete for a huge amount of premium digital content. This isn't to say the US dollar is obsolete, because you can represent dollars, or euros, or gold, whatever, as auxiliary currency units on top of the Bitcoin blockchain, although you'd still need to own BTC to send it around, for many this is BTC's primary form of inherent value.
Finally, it's impossible for many people in developing nations to engage in online commerce at all, because their banking system is utter crap. Billions of people the world over are underserved by the banking system, or don't have access to a bank account.
That's just silly. You can certainly think of things people have a legal right to do that they still wouldn't want any record of. If you're paying for an abortion it's perfectly reasonable to want the payment to be truly anonymous. The same goes for other types of medical treatment, any purchases of an embarrassing nature, donations to politically charged organizations like Planned Parenthood, charity by people who believe altruism is only satisfied when the donation absolutely cannot be attributed, etc. There is also the simple metadata concern that giving anyone a record of everything ever purchased by you (and everyone else) could be used against you or for evil purposes generally.
The ability to make anonymous payments has been the norm since money has existed. Cash is anonymous. It's only the last few decades with the rise of electronic payments that tracking them all has even been possible, and there is no law of the universe that says we have to accept the loss of privacy.
I think most people would welcome a shift to more anonymous transactions, but they're not willing to give up the advantages they get from tracking. When my purchase is tracked by a the seller, I get volume and loyalty discounts and easy returns. When my purchase is tracked by a credit card company, I get the ability to dispute payments and better manage cash flow. Most people and businesses aren't willing to give up those conveniences for anonymity.
The ultimate reason why I have a bank account is so that my employers can deposit their payroll into it. And since my money is already there, I pay my bills and expenses from it.
I fully expect my bank to narc on me to the feds at every turn, and screw me hard if I ever have a real problem with them. I expect them to nickel-and-dime me with fees. I expect them to introduce seemingly arbitrary delays and paperwork requirements for their own convenience. I expect them to offer rates on deposits that cannot keep up with inflation and charge rates on loans and credit that would make even the serenest of divine avatars flip tables. And if the turds ever hit the turbines, I expect that making a trip to the nearest bank branch office would be the least useful thing I could possibly do.
I don't really have a lot of confidence in Bitcoin, but the bar they need to pass with respect to trust is ridiculously low--set there by the banking system. Where they fail is definitely convenience.
If my employer offered to make payroll in Bitcoin, and if my local utilities, gas stations, and grocery stores accepted payment in Bitcoin, I would be using it already.
But all those places prefer to trust human bankers rather than crypto algorithms running in a distributed computing network. I don't see that changing any time soon. If anything, an alt-currency backed by an alliance of megaretailers like Walmart, Target, and Amazon would have to crack that nut first. To drive adoption, there has to be support at the point-of-sale terminals.
My feeling is that the interesting thing about Bitcoin as a currency is the pressure it puts on centralized payment systems to compete with it. The problem with banking is that it's so hard to innovate, primarily because of the regulatory burden both making it very expensive for anyone to enter the market and severely constraining what can be improved if they do.
For example, with credit cards the burden of fraud is put on the credit card company. That turns out to be one of their major costs, even though the credit card company is hardly the lowest cost avoider for credit card fraud (they're just a convenient deep pocket), which requires them to charge high transaction fees to cover the liability. Hence no micropayments; fees too high.
Now suppose that Bitcoin disabuses people of the notion that the payment processor should be the one who takes the hit for fraud. If you don't secure your secrets then you lose your money; you can buy insurance if you like. Suppose regulations get passed to allow this for Bitcoin. Now non-Bitcoin payment processors can make the argument that they should be able to offer a centralized service with the same rules, and we can finally have micropayments.
[1] https://www.wto.org/english/Tratop_e/dispu_e/cases_e/ds285_e...
What about skype, that was built upon supernodes, and was decentralised?
Skype accounts were centrally registered and a single company was responsible for network health, security, etc.
>(Netflix et al) are far more popular and commercially successful for watching online video
File sharing versions are not aimed at being commercially successful so i think that is an irrelevant comparison.
Also you I expect are from the USA, netflix/hulu etc arenot available everywhere so they are certainly not more popular than the alternatives in countries where they are not avaiable.
Popcorn Time was a free, decentralized Netflix alternative and it was hugely popular, only that the developers pulled the plug on it did it not become the defacto method of watching movies for free online.
Anyway not sure how these compare to bitcoin but I dont think your analogy holds up even when not compared to bitcoin.
When there are so many similar versions which one do you choose.
Also you probably don't want it to be popular anyway unless you don't mind the authorities knocking on your door.
I think the market will always decide which media consumption method it wants to use depending on how frictionless the experience is.
In the case of Bittorrent, as a mainstream consumer you'd have to know what program to download, where to find a torrent file yada, yada yada, while on Netflix you can just enter your credit card information and watch as many movies as you like.
I think recently, Facebook also realized this, which is why they are currently aiming towards bringing a great search engine to their customers.
Once that is in place sufficiently, homepages for small businesses will become obsolete, as people rather search for a company/product on Facebook than on the internet as it is a much easier to do.
This same principle again will probably apply to Bitcoin and digital payments in general. Right now, Apple is doing a great job reducing friction in payments, at least for mainstream consumers.
While this may or may not be true, am I the only one who is slightly frightened by the ramifications of this?
Facebook is fine, for what it is, but the idea that it could soon become a proxy for the WWW isn't my idea of optimum.
But I think the only reason Netflix doesn't use BT is that the content owners are super-skittish about people having a copy of the data on their computers instead of streaming it and throwing it away. So even if BT is a better technical solution, it's politically unworkable.
It indicates that BT didn't provide something that Netflix et al did require. I imagine that is something like maintaining control of all digital copies of a work as you went on to suggest. This is not a technical limitation but rather a business need. So BT may be technically better (I imagine in terms of resources required to provide content it is superior) however it didn't meet the business case as well as a centralised solution did hence BT doesn't get used.
But consider for a moment the popularity of Popcorn time or torrents in general. They're massively popular despite the fact everyone knows it's illegal and knows many countries have histories of punishing illegal downloaders, let alone the big uploaders. And it's not just that they're popular, they work really, really well for something completely free.
Does that mean we prefer using commercial solutions most of the time? It depends. Torrenting tech is monetized through ads but kept free, just like things like gmail or facebook. The difference is that gmail isn't illegal and thus can get venture capital investments, partnerships, build elaborate products (like gmail for business) that interface with online identities (like businesses, phone security) and infrastructures (banking & payment processing infrastructure), have customer service etc. While most torrenting services have to operate in the dark, anonymously, outside of infrastructure, with few partnerships, little investment etc.
In short, I don't think we should try to reason from analogy here. Centralized/decentralized doesn't inherently predict anything.
I think at the end of the day we have to look at the 'per person' scale. The average US citizen makes 2 transactions a day (one of the metrics I saw used in this slideshow, too.) The average bitcoin transaction is 500 bytes. For a global citizen it's likely slightly less than 2 on average considering the US one of the most consumerist societies, but let's say 2 for everyone.
Alright so 1 kilobyte of data per day for everyone in the world.
Is that a big deal? No. It's not. It's still 14 TB of data, daily, but it's manageable.
It requires an 82 mb/s connection to handle that much data. That's for the entire world populations (1 month old babies included!) to average as many transactions as the average US citizen.
Just put that into perspective. We're not talking about this kind of technology today, bitcoin has only a few million users. It's unlikely bitcoin will ever become popular enough to service 7 billion people in the first place, the internet's been around for decades and it's not even close, hell electricity or clean water doesn't even do that. But hypothetically if it happened, we'd be talking about some date that's at LEAST 20 years out from now to get to 7 billion and that in itself is already extremely optimistic. For perspective, it took the internet nearly 2 decades to get to 2 billion internet users.
So to say a 82 mb/s internet connection in 2035 is too big of a deal for what is likely going to be a node run by a professional business like Coinbase. (i.e. I believe bitcoin's infrastructure will be run more and more by professional businesses rather than home desktop PCs, but these businesses will exist in every country on the planet and be decentralized a bit like how ISPs are decentralized only much more because running a node isn't nearly as big a deal obviously, as it's one of many things running on the ISP's network).
Average home internet is about 5 megabytes per second today in OECD countries. And we know bandwidth, like CPU and storage, has grown roughly 50-60% in the past decades and it's likely to average something close to that (like 40-50%) in the next few decades. 50% yoy would be 3300x in 20 years. 5 MB/S would be 15 gb/s by 2035. Optimistic? Absolutely, it's the most optimistic metric based on past data. But let's just say it's extremely likely that reaching 1 gigabyte per second will easily happen by the time a bitcoin node needs 0.082 gigabytes per second. And that's for a standard, average, consumer, home connection. NOT the much better industrial connections that you expect bitcoin nodes will be hooked into by 2035 (if it indeed is so important that 7 billion people use it for 2 transactions a day).
In short, 1 kilobyte of data per person is so little that whether it's centralized or decentralized it doesn't really matter that much. It can be dealt with. Especially when you consider that many of these transactions can happen off-chain, on sidechains, treechains, less secure chains (e.g. to buy coffee or make a 1 cent payment to read a New York Times article automatically instead of being served a 10 second advertisement with a 1 cent revenue per view value, these transactions can happen on less secure channels).
Decentralized/centralized is an interesting discussion, sure, but at the 1 kilobyte per day per person scale, I don't think it's very relevant when it comes to technical scalability. It's relevant for other reasons definitely (like a decentralized digital global banking system being much less prone to anti-competitive banking monopolies and licensing than say a centralized one like Paypal) but I don't think technical scalability is the biggest problem.
Beyond that bitcoin is so nascent and small... I mean I quite like it and think it just might win, but I still feel it's a bit silly to talk about whether bitcoin can scale to a billion users all the time, when its growth rates are really not all that impressive. I mean don't get me wrong, scalability is important to talk about, but it commands such a disproportionate amount of talk. It's like wondering if 50% of the Mars debate was about whether Mars can support a city of a million humans, let's see if we can get 20 people living there first you know?
I would disagree that the reason people use Netflix instead of bittorrent is not because Netflix is "less work" or that "there are costs to be paid by being decentralized". Your argument above is ignoring the legal issues with transmitting copyrighted data that deters ordinary citizens concerned about following the law. I would actually say that Netflix is more work than bittorret, espcially compared to the bittorrent clients nowadays that can stream bitorrented video. Plus in the past netflix always required a bloated plugin (silverlight).
The Gnutella with super hubs was called G2 protocol. Gnutella also utilized GWebCaches for retrieving bootstrap information of other active nodes.
This seems to be overly optimistic considering that the acceptable credit card chargeback rate is 1%. 36 times per the rate of channel roll overs.
In addition even nicely working channel has to be created. So even without anything like chargebacks assuming 3650 transactions per year on a single chain seems to be extremely optimistic.
It might reduce spam on the network though. Bringing us from the 240G blocks into 24G blocks (Based on the well known Stetson-Harrison method). A fine improvement, but doesn't really seem to be a silver bullet for the scalability issues.
That may be true. But even when multiplied with your proposed factor 36 it still looks very promising:
~4.7G blocks, ~108Mbit/s, 252T/yr for full blockchain archival.
In fact, that looks almost too good to be true for a blockchain with 7 billion participants. Did I make a mistake in my math?
Keep in mind that (if I understand it correctly) these are only the requirements for miners ("supernodes"). Mere mortals would largely operate off-chain (and SPV) with near-zero bandwidth and storage requirements.
So even in the best case it would be "Instead of scaling ridiculously badly bitcoin scales only horrendously badly".
A natural solution for this, if we really really want to get cryptocurrencies to work, would be to basically have a cryptocurrency per nation or so, with different PoW function to avoid shifting asics around. The only reason Bitcoiners want a person in Argentina to keep a Finnish potato sack purchase in their books is just to pump up the value of their 'investment'. Every solution made in relation to Bitcoin is made from the viewpoint that if it would make any current Bitcoin holders potentially miss out on global domination it's ignored. Even if it would actually work better.
Huh?
The above numbers are well within the reach of most western individuals even today. A rented server in the $50/mo range can trivially plow through ~100Mbit/s of SHA256 verifications.
250T of storage on spinning rust can be rented for ~$1000/mo from e.g. Hetzner (and obviously a fraction of that if you roll your own).
By the time a Bitcoin adoption of 7 billion users becomes anywhere near realistic you will probably have this kind of storage and CPU capacity in your Smartwatch.
If this type of processing power became true, other things would become untrue. For example the cumulative hash-breaking power of the world would be much, much higher. Also we'd have effectively limitless energy sources (how are you powering that multi-core smartwatch?)
And it would still be a much better use of resources to do literally anything else for currency.
Why would the "cumulative hash-breaking power" grow somehow disproportionally to the general computing power?
how are you powering that multi-core smartwatch?
Likely in the same way as your current multi-core smartphone?
Of course, this is already true anyway - the largest mining pool has been large enough to carry out 51% attacks for a while now, even before it was shown you could do it with much less power.
>Likely in the same way as your current multi-core smartphone?
Battery technology hasn't gotten appreciably better in a decade. And even if you could get enough power, how do you handle the heat? And even assuming all that...you then run into the problem above. The power of every other computer in the world will be a lot higher.
"Could miners fundamentally change the nature of Bitcoin?
Once again, almost certainly not.
Bitcoin is a distributed network, so any changes implemented to the system must be accepted by all users. Someone trying to change the way Bitcoins are generated would have to convince every user to download and use their software – so the only changes that would go through are those that would be equally benefit all users.
And thus, it is more or less impossible for anyone to change the function of Bitcoin to their advantage. If users don't like the changes, they won't adopt them, whereas if users do like them, then these will help everyone equally. Of course, one can conceive of a situation where someone manages to get a change pushed through that provides them with an advantage that no one notices, but given that Bitcoin is structurally relatively simple, it is unlikely that any major changes will go through without someone noticing first.
The fact that such changes are so difficult to make testifies to the fully distributed nature of Bitcoin. Any centrally controlled currency can be modified by its central agency without the consent of its adherents. Bitcoin has no central authority, so it changes only at the behest of the whole community. Bitcoins development represents a kind of collective evolution; the first of its kind among currencies."
the Bitcoin hash function doesn't change, can't change
Of course it can be changed with a client update, like everything else about bitcoin.
SHA256 is getting weaker proportional to all power in the world
I'm not even sure what that is supposed to mean...
the largest mining pool has been large enough to carry out 51% attacks for a while now
There is currently no pool anywhere near 51%; https://blockchain.info/pools
Battery technology hasn't gotten appreciably better in a decade
Li-Ion battery capacity has been improved by 20% last year; http://en.wikipedia.org/wiki/Lithium-ion_battery#From_commer...
Each party to the side channel has a multisig transaction signed by the other parties involved, so if one party cheats (or fails to comply for innocuous reasons), then any party can shift the entire chain of transactions to the blockchain without risk. So the only reason you don't want side channels to persist for too long is that each party has to keep around more and more information as the number of transactions over the channel increases; that's burdensome in its own right, but it also likely means that a side channel naturally decays over time, as the probability increases that one of the parties involved drops out or loses track of information. So you don't want to roll over side channels too many times, but it's not a matter of risk of trust.
This really is a very simple and elegant solution. Basically, it recognizes (1) that cryptographic verification of intermediate transactions only needs to be shared among the parties involved, rather than among the entire network, and (2) that any sequence of transactions among a sub-network of parties can be considered as a string of intermediate transactions. And (3) it requires no trust because the entire sequence of transactions can be moved onto the blockchain by any party at any time.
Consider the amount of transactions you do. Basically every party becomes a single side channel which requires some activity in the actual blockchain. It would reduce the amount of transactions that are the style of "My local supermarket which I visit every other day" but not anything else.
Combined with increases in the block-size limit, this technology and others are increasingly making it seem like scalability is unlikely ever to be a real problem for the Bitcoin network. (And file that under great problems to have.) Recently I've seen scalability cited as a demerit of the Bitcoin network more and more by people who don't keep up with the technology and the community developing it—and less and less by people who do. Good progress is being made.
But I agree with you that the rough assumptions the authors used to quantify this technique's potential to scale the network don't make much sense.
Pros:
1. They already have a paper and I don't :-)
2. No need for extra capital lock up between nodes to cover bidirectional IOUs.
Cons:
1. Channels can be "closed" any time forcing everyone in the chain (could be 7-10 nodes) to make a BTC tx whether they really want it or not.
2. Necessary timeouts require channels to be closed after some time, while in theory some of them can be kept open indefinitely. In my scheme, interior pairs of nodes can pay each other accumulated transaction fees in separate transactions whenever they want without affecting anyone.
As for people who do need the full blockchain in the future scenario you describe, they could always pay someone to sneakernet it to them.
I just imagined an armed convoy protecting a truck loaded with hard drives.
Bitcoin core latest version (0.10.0) have reduced significantly the bootstrap time of a node (e.g. through "header-first" mechanism).
Future releases will introduce other improvements that will lead to a reduction in terms of network bandwidth/latency (O(1) block propagation, IBLT) and IO space (e.g. pruning)
That said it's true there's no economic incentives in place to "just" run a full node. Person that run full nodes does that for a lot of reasons modulo the economic one.
Hence the number of full nodes is steadily declining http://www.bitcoinpulse.com/#/chart/bitnodes/num_nodes
There are proposals (1) floating around to fix the situation through, guess what... economic incentives, but we are for from getting some sort of agreement, worse than that I dare say that in this case we're still struggling with problem awareness.
(1) https://bitcoinism.liberty.me/2015/02/09/economic-fallacies-...
That's really not a problem, centralized or decentralized, for bandwidth, storage or cpu to handle.
Fact of the matter is that a transaction costs some money, like 5 cents, which is in a way the price to transmit 500 bytes of data, verify it and store it. That's economical at any scale. And guess what, the marginal cost is only decreasing due to technology improvements. (moore's, kryder's and nielsen's laws)
The price of 1 gigabyte of storage is about 3 pennies today (retail), for example. And storage is not permanent either, you can prune old transactions (that's an innovation currently in development).
And again, that price is only dropping, in 1990 that price was a few thousand dollars by comparison.
:-/
Check out this image:
http://cffn.ca/img/articles/Centralized-Decentralized-And-Di...
Bitcoin is both distributed and decentralized. In a world with 100 people, you can say that a system is sufficiently 'decentralized' if 10 of the people ran nodes, independent of eachother, where if any of the nodes lied or crashed the system would continue to work, and where few or no nodes collude with each other. It doesn't require every single one of the 100 people to hold the blockchain.
Bitcoin nodes is a bit like peer review of a dozen academics keeping each other in check, only it's based on math and cryptography and proof of work. The proof of work ensures it's not possible to lie to other nodes without expending lots of money on mining and risking not getting any reward in return when your blocks are rejected. And even if you lie, nodes can't send money they don't earn due to the public-key cryptography. And instead there are hundreds or even thousands of such nodes and miners, and we call that decentralized.
In other words you can run a decentralized system with thousands of copies of all transactions, as opposed to all 7 billion people needing to store a copy and relay every transaction to everyone else.
Beyond that, not all transactions need to even be stored by the nodes that are needed. If you sent me a dollar in 1990 and I spent that dollar, and whoever received it from me spent it and so on tens of thousands of times now in the past 25 years, then there's absolutely no practical reason for me to store the data on our transaction in 1990. It can be pruned. And so can all the transactions between 1990 and 2014, and most transactions in 2015 of that particular dollar, except the last recent few.
Concerning security of the blockchain the vast majority (an ever growing number) of transaction data can be safely thrown away.
Some research universities and businesses might save that data for posterity/research/datamining, but for the network to operate, to function, you can prune something like 99% of the data.
The Bitcoin transaction rate is currently only about 100K/day, or 1.1/sec. That's only twice what it was two years ago. There's no urgency in solving this problem.
At ~$30,000 a year in credit card spending I'm throwing that amount away in credit card fees every two and a half years. I'm perfectly capable of and would prefer the option to self-insure, thank you very much.
Ultimately insurance always has a negative expected value for the average consumer. It makes sense to pay the vig to insure against catastrophic events like a car crash or a house fire, but insuring my daily credit card purchases like takeout meals, groceries, and gas is stupid.
However... bitcoin does allow new insurance providers to pop up - the insurance provider no longer has to be the same entity that holds your cash (your bank) or provides payment processing services. They could experiment with new models - multi-key escrow until service delivered, an "approved store" model where they'll provide insurance on purchases from stores that they've checked out without ever having to touch the money themselves, or more. This could all be cheaper than what your processor + bank combination costs, potentially.
The thing is... that's not really good enough for customers. Nobody thinks enough about a few pennies per purchase to pick up a whole new currency.
Bitcoin beats "Cash On Delivery" because it is trustless, but that's easy to address by having the delivery service allow a remit prior to the shipment.
Note that with CBD, a (cash using) recipient has to deal with 1 party, with bitcoin they have to deal with 2 or 3 (obtain, spend, receive).
Services would be somewhat different, but it still seems like quite an open question whether bitcoin would (in practice) be better than buying some credit from a payment provider (Paypal or Amazon or whatever).
But you could have said the same thing about email/the web in the past.
A global communication network and protocol without anyone using it is pointless. The same for a global value transfer protocol. It seems obvious _a_ protocol, probably blockchain based, will gain mass adoption eventually.
I also doubt that people who don't have bank accounts have private computers. So really the pool is "people with computers and bank accounts in countries so replete with fraud that Paypal and CC companies don't service them".
The argument is that they'd be paid in Bitcoin for goods/services in the first place. It is a currency, after all, just (perhaps currently, perhaps always) a very volatile one.
With the Lightning Network, you could transport any amount of value, instantly, for very close to zero cost, allowing things like paying a random untrusted party for WiFi by the second, with incremental micropayments.
Only in a fictional world where my employer/clients pay me in bitcoin.
In reality I need to go to an exchange, convert to bitcoin, then go back to either the same exchange or a different exchange and convert from bitcoin, all in all I've gone through two intermediaries and paid two fees.
For one example if I did that on CoinMill they would charge me $2.59 over market for 1 bitcoin ($252~254 in value) and the same again when I convert it back to dollars. $5.18, that's approx. a 2% transaction fee. TransferWise by contrast (who really are super hassle free) charge $2.88 for one end-to-end transaction, so less than bitcoin(!).
So yeah, your argument is completely broken in the real world. Only in a fictional world where everyone uses bitcoin does it work.
Why "everyone"? Wouldn't it be enough if a shop that has something that you want to buy accepts it?
There's a great many shops accepting Bitcoin today already.
Certainly doesn't result in: > You can transport value across the world instantly, with no intermediaries. This is a major efficiency gain.
Since (A) there is an intermediary and (B) it is not efficiency (which I'm assuming is a euphemism for cheap anyway).
There's scenarios where TransferWise, MoneyGram, WesternUnion etc. either don't work or ask for extremely high fees.
It can also be simply inconvenient to go to the physical pick up locations of MG/WU. Keep in mind most people in the world are not blessed with 3 different banks and a starbucks within walking distance.
The intermediaries are domestic exchanges in this case, meaning only domestic transfers of fiat need to be made. The international transfer occurs through Bitcoin. Fees for large Bitcoin exchanges range from 0.25% (Bitstamp) to 0.38% (BTCChina). Coinbase currently has no fees, but it's a temporary promotion. That means it's less than 0.8% in trading fees altogether, when using major exchanges. Assuming Bitcoin becomes a little more widely adopted, this situation will only get better.
I think the real opportunity is in the back end of the financial system. Banks can begin to use Bitcoin for international wire transfers, while enjoying much lower trading fees than retail clients when converting into and out of fiat currencies.
[0]http://www.coindesk.com/companies-paying-employees-bitcoin/
Do you have an idea of the quantity of power Chinese bitcoin mining farms consume? That's only for the few transactions that the networks processes right now.
Here is a quick computation that shows that every transaction that goes through the Bitcoin network is equivalent to releasing 47kg of CO2 into the atmosphere: http://redd.it/2hiea4.
Other than the accelerating effect of transaction fees on mining resource consumption, which is on the order of 2 cents per tx, there is no relationship between number of transactions that the network processes and the power Bitcoin miners consume. In other words, the cost (both economic and environmental) per transaction will decline as transaction volumes increase.
The current Bitcoin network cannot handle that many (think an order of magnitude) more transactions without having to scale. See also the original link.
Almost the entire cost of mining is in running ASICs to produce proof of work, not in validating transactions. The cost to produce proof of work is not affected by number of transactions.
The SQL Database Model ○ Very scalable, very fast