This is basically saying "Bitcoin might not work because some nodes might decide to go start another currency."
That's a threat to any monetary system. With Bitcoin, as soon as nodes start breaking the rules, they are ejected from the system. The fact that a bunch of them could leave together and start a new system can hardly be considered a real criticism of the protocol itself.
It isn't a criticism of the protocol. It's a threat to the current system. A centrally-controlled, government-sanctioned/mandated fork of the protocol that removes the bulk of the desirable properties could starve the "good" BitCoin by stealing away the vast bulk of the users that BitCoin needs for long-term effectiveness, and long-term protection from the charge that it is only useful for money laundering and drug deals. That charge needs to be visibly false for it to stand a chance.
The best way to guarantee that this threat manifests is to just poo-poo it as an impossibility, and not prepare for it. No matter how you slice it, BitCoin faces an uphill battle to get to the point where it has enough acceptance to dodge the laundering charge, and a quasi-religious denial of the problems because "BitCoin is just too awesome to be affected by anything bad!" is a quick path to total failure.
One entity dominating the network may not be a serious threat. A cartel damn well is. The article is correct, the resources are not centrally distributed and you don't need anything like 50% of the humans to cooperate to end up possession 50%+ of the network power, especially if you've got the men-with-guns backing you and even modest cleverness.
I've seen a couple of dismissals of this possibility, linked elsewhere in this comment section, but it's like the BitCoin proponents suddenly forget they're up against men with guns, despite starting out with this understanding, not just some people who want to play modestly unfairly and merely "collaborate when they shouldn't", but otherwise follow the rules. Effort bent towards breaking BitCoins will be proportionate to the size of the BitCoin economy.
A "centrally-controlled, government-sanctioned/mandated fork of the protocol" would steal few users from the BitCoin economy, because the kind of people that would move to the new system were probably using fiat currencies to begin with.
Men-with-guns haven't brought down p2p file sharing yet.
> Men-with-guns haven't brought down p2p file sharing yet.
But they have brought down or kept down every single digital currency to date, save for those that were merely scams to begin with.
P2P file sharing is not perceived as a threat to the revenue model of government, and is not especially attractive to mobsters and scammers. P2P currencies are.
I'm not a libertarian, yet I like BitCoin better than fiat currencies and centrally manipulated monetary systems.
If other people are happy with whatever features of BitCoin remain under centralized control, more power to them. I see no problem.
But you're saying we have to fear that someone will start effectively issuing p2p dollars. I think it kind of defeats the purpose, but I can't see why that's a more of a threat to the principles of Bitcoin than plain old USD.
Because of loss of value. Without backing there would be nothing guaranteeing the BitCoins in your wallet would hold close to the same purchasing power from before the time the network split.
It would not be so much a network split as a network copy. Every user of the original network would have, to begin with, identical balances in both. Until inflation adjusted things, there would be an ephemeral increase in purchasing power.
If the new network implemented inflationary logic, as a modern state is wont to do, I can't exactly see people rushing to convert their old (relatively stable, perhaps deflationary) money to the new one. Therefore, I can't see many businesses stopping to accept the old bitcoin.
And after whatever adjustments are made, it will be business as usual. Bitcoin will remain the network for people that want a decentralized currency. Not necessarily of libertarians, but of anyone with distrust or disillusion with their governments and the banking system. Which I don't think is such a small demographic.
I can't exactly see people rushing to convert their old (relatively stable, perhaps deflationary) money to the new one. Therefore, I can't see many businesses stopping to accept the old bitcoin.
Rushing to convert "old network" BitCoins into the new network (or dollars, or whatever), and businesses stopping acceptance of them is exactly what will happen. Don't forget the primary function of money is to allow people to trade it for things they want. Unless you're a staunch ideological libertarian you will want as much of your "money" as you can get in whatever system offers the most goods and services. Otherwise you have lost value. That system will certainly not be the one containing only sparsely geographically spaced libertarians.
We already have "A centrally-controlled, government-sanctioned/mandated fork of the protocol". It's called the federal reserve system and it is very networked, efficient, etc. and has almost none of the features of bitcoin because it doesn't need them. All the features of bitcoin are there so it isn't possible for any one entity to centrally control it. It would be ridiculous for a fork to be created that would be centrally controlled. It would be like saying someone should fork open office and turn it into a submarine control system.
How would this centrally-controlled, government-sanctioned fork differ from existing fiat currencies like the dollar? If we assume that a large number of people start using Bitcoins, why would this network be threatened by a competing currency that has no advantage over existing currencies?
I think this article nails the description of a problem that Russ Roberts also alluded to on EconTalk -- that bitcoin can change and those changes are made somewhat democratically.(http://www.econtalk.org/archives/2011/04/andresen_on_bit.htm...). I think libertarians assume that the bitcoin network will be full of other libertarians. The article is right that the network may also be full of big entities whose behavior is regulated. Whatever your feelings about regulation, this would seem to make bitcoin less interesting.
I think one option is to accept that the bitcoin protocol will be forked massively and design it from the ground up to support exchanges between various versions of the protocol. That way, libertarians could avoid versions with inflationary policies. In this setup, would there be any reason for an entity to choose an inflatable flavor of bitcoin? Perhaps certain businesses and banks would be coerced to through regulation and that would create more widespread demand for those forks. It seems possible that most users would use the regulated bitcoins and the libertarian version might become a black-market currency.
All this feels not too different from the current situation where you can invest in gold or leave your money in a fiat currency. I guess the main difference is that the "gold" (i.e., the non-inflating bitcoin forks) might be less traceable and more liquid than true gold is today. I'm beginning to suspect that, while bitcoin is a damned-cool technical idea, it isn't likely to massively change the world other than, perhaps, to grease the wheels of black-market transactions.
>I think one option is to accept that the bitcoin protocol will be forked massively and design it from the ground up to support exchanges between various versions of the protocol. That way, libertarians could avoid versions with inflationary policies. In this setup, would there be any reason for an entity to choose an inflatable flavor of bitcoin? Perhaps certain businesses and banks would be coerced to through regulation and that would create more widespread demand for those forks. It seems possible that most users would use the regulated bitcoins and the libertarian version might become a black-market currency.
And to do this, that means that you need some trusted way to determine the price of Bitcoin 1.0 versus Bitcoin 2.0.
Which means you need a centralized place, let's call it an exchange, to determine the rates between the two version. Since you probably want the rates to be consistent across the entire network at any given time(otherwise, what's the point of a non-inflationary currency), it's probably a decent idea to maintain a centralized location for that that decides the rates. A decentralized method may be possible, but I sincerely doubt that it will work in a way that would actually encourage people to convert to the other currency to purchase things.
In either case, you still need something that can regulate the transactions to make sure that everyone is behaving correctly. If people aren't(which should be the first assumption you make about anything), you'll need some method of dealing with it.
So... yeah. You just created the Federal Reserve and money markets.
Realistically, given that it's a democratically run, the current Bitcoin implementation is in the same boat, if the biggest users decide to change their policies, the entire network can end up in a similar situation where a Bitcoin from X is worth a different amount from a Bitcoin from Y.
I don't really understand how Bitcoin is better than the current currency system. It's just different because it's on the internet.
A decentralized version is virtually guaranteed in the absence of a centralized one. If you have many exchanges, and they offer different rates, arbitrageurs will exploit this and make the rates match.
This is why, for example, the price of RIM is virtually identical on both TSX and NASDAQ.
This is exactly what I thought when I saw all of the bitcoin articles on HN last week: essentially, the people who control the top Bitcoin applications are like the central banks: They control the protocols/how much money is released and also how Bitcoins are mined.
The Bitcoin protocol fixes the rate of inflation by adjusting the size of the hash that needs to be calculated. Big Bitcoin rigs don't cause inflation, they merely increase the odds that newly minted Bitcoins will accrue to them.
Look at it this way: It would be approximately as hard to change the rules of a mature bitcoin network as it would be to change say, the rules of the web (perhaps adding something easier to implement than <video>; more of a server-side change like say, TCP fast start) ...
But to make the change stick in the bitcoin network, the change would have to be made to all peers, network-wide, in the time it takes to calculate a few blocks. Otherwise the fork doesn't succeed.
The rules of the web are changing constantly. All you have to do to change the web is convince Google, Apple, Microsoft, and Mozilla to go along with it.
Not only that, but there are fundamental limitations of the protocol as it exists right now which will cause BitCoin to fail if it does not change. For example, blocks can only contain 1 MB worth of transactions. That works out to ~3,000 transactions every ten minutes, or about 5 per second. By comparison, Visa does about 8,000/second.
OK, here's plan B. You just start a new network and seed it with a copy of the old block chain (to bribe old users into at least trying it out). No need for elaborate and potentially expensive takeovers. Then you advertise the new network as the one sanctioned by Your Friendly Government.
What does the OP's elaborate, hostile, expensive plan buy you over plan B? That users will have to opt-out of your branch, rather than opt-in? Is that wise PR?
Although this is true as I understand Bitcoins it kind of ignores how it would work in practice.
Essentially, if we had a mature Bitcoin network, then we'd all have some money in it. And we obviously have an interest in keeping that money valid.
Sure, half the people could fork off with new rules for Bitcoins, but they leave behind everybody's old bitcoins in the process. As others have commented, it is a fork of the currency, but you don't get to keep your old bitcoins.
So although the rules can change, you start from scratch again. Your old bitcoins won't work in the new world of 100 bitcoins / 10 mins, and those new bitcoins won't work in the old world.
So we'd all have to start over. Ergo, it doesn't seem very likely.
If they fork the block chains, my understanding is that your old bitcoins will work alright in either network. People that stick to the old network won't accept any money from the new one, but nothing keeps them from spending their money in the new one. Thus the adopters of the new economy would be subsidizing the members of the old one.
[Edit] And now my own semi-rebuttal: sure enough, the people who move to the new economy can similarly spend their old credit in the old one. But considering that the original Bitcoin currency is deflationary in nature, they have less incentive to do so, at least immediately.
Ah, so you are saying that I can use the same coin twice: once on each network. Basically, as long as the networks keep splitting the coins can be re-used. However, this would devalue the coin, independently of whether the new networks introduce further inflation. This seems undesirable to most people on all the networks: if by splitting I can still only buy 10 cups of coffee with my bitcoins, just now it's more of a pain because I have to remember through which network to use it, I wouldn't want to do it.
If someone started a new Bitcoin-like currency they should start from a blank slate. Cloning the network chains would be confusing and frivolous at best.
I would really like to see a conversation between Timothy and Jonathan about this; it seemed like they were the only ones with useful comments on that article, and everyone else missed the point Timothy was making.
Reading through those comments to follow the interesting parts of the discussion made me depressed.
90% of the commenters failed a pretty basic reading comprehension test, maybe I shouldn't be surprised (it's the internet) but I figured a bitcoin related blog conversation would be substantially different from youtube comments.
The other threat to such a system, when it gets larger, is that others who make interfaces to store/use BitCoins might also allows other BitCoin-esque currencies. Which is fine if each of these currencies is vetted the way BitCoin is undergoing right now, however if it's very automated we could see new distributed currencies issued when the demand for new BitCoins is high enough.
TL;DR: Other BitCoin-like currencies could be used to grow the money supply - which might lead to bubbles of such currency or perhaps solve the growth issues BitCoin will deal with.
I love the idea of alternate currencies and think BitCoin is a step in the right direction but this article points out the obvious -- the # of bitcoins does not grow is a straight up lie. I don't know why people keep believing this when you can look at the design and plainly see that it's not true.
The published spec conceives of halving the number of bitcoins a block miner receives every n years (4 or so?). Right now, a block miner will get 50 BTC. In a year or two, they'll only get 25. A few years later, 12.5. The Sum of the BTC generated before the block generation gets really really little is roughly 21 million.
As has been pointed out copiously today, that's a convention that BTC clients and users all claim to agree to. It seems likely to me that they will continue to agree to it, and it also seems likely to me that the major hurdle will be the first '25 BTC/block' downgrade. If the community survives that, the others will likely follow easily.
The feasibility of a challenge to any reigning monetary scheme creates great incentives to bolster its integrity. Lacking integrity, the last resort is of course coercion.
Already discussed months ago. Bitcoin critics are for some reason less knowledgeable about potential weakness of bitcoin than bitcoiners who are enthusiastic about bitcoin.
Yes, but in all fairness, that post addresses problems that can result from a computational attack. The one you raise doesn't need such at all. If a state wanted to compete with Bitcoin on all features but distributedness, a peaceful fork in an independent network would have essentially the same effect.
If P = NP, that doesn't mean there are practical algorithms in P for the problems in NP. It could turn out that for all the problems in NP that we currently think aren't in P, the solutions in P are O(n^100000).
Also, it can go the other way. Just because a problem is NP complete doesn't mean that if P != NP we can't solve it efficiently in practice. It just means we can't get an efficient algorithm that will work in every single case. It may turn out that the problem instances that actually come up in real life are amenable to fast solutions.
But I'd like it much better if the massive amount of computing power being poured into the hashes was actually being used for a good cause like folding, seti, etc. and the bitcoins were still the reward
(new) bitcoins can't be the reward of a project like seti@home, because to find a valid bitcoin in the first place (to mine it) takes that much computing power. Now that some valid bitcoins have been discovered, those cryptographically valid 'coins' (hashes) can be used as money. Such as a reward to SETI@home users. But whoever takes the time to mine that unique, distributed blockchain (the consensus' transaction log of all bitcoins) has to choose to give the SETI project those unique bitcoins. The process of rewarding a good deed and the process of mining a safe blockchain are independent, and can easily take many thousands of volunteer computers, respectively.
If you can design a system that generates a cryptographic secure chain of the collective transactions of volunteers computing unrelated problems... you should publish the design.
That's a threat to any monetary system. With Bitcoin, as soon as nodes start breaking the rules, they are ejected from the system. The fact that a bunch of them could leave together and start a new system can hardly be considered a real criticism of the protocol itself.