What's not to trust? It seems that the clients determine what they trust, and so long as the majority of clients agree, all is well.
There are economic problems with bitcoin as a currency (its a fiat currency, effectively, except without the fiat - it has scarcity with no intrinsic value), but the algorithm component seems solid to me.
All the fiat offers is confidence that you won't be the one left holding the hot potato. Seems to me bitcoin provides the same confidence through mathematical cleverness, which is actually more trustworthy in the long run.
Bitcoin doesn't offer that confidence, though. No currency does, fiat or otherwise. The confidence in a currency comes entirely from the economy that develops around the currency. In other words, the power of a currency comes from what it can be used for, not it's scarcity. If you cannot use Bitcoins, USD, or any other currency, they are worthless.
For example, if an overwhelming majority of people with bitcoins(or USD) decides tomorrow to not ever use them again, then the remainder that have them are screwed. If the places that the currency can be uses cease to exist, the currency is worthless. There isn't any technical solution to this problem, the economy just ceases to exist. Obviously, this is an extreme case, but it's entirely possible that this can happen with any currency over an extended period of time if it doesn't catch on, or if investors lose confidence in the currency itself.
Sure. Certainly. But that is true of any currency. The novelty in bitcoin is that it enforces scarcity, and thus limits inflation, without depending on a central issuing agent.
Part of what it means to have confidence in a currency is to be reasonably sure that its inflation will be limited to a rate that allows you to exchange your currency tokens before they are drained of value. Passing the hot potato, etc. With a fiat currency that confidence is based on nothing more than your beliefs about the central issuing agent's likely future actions.
What's new with bitcoin is that there is no central authority who can decide what the inflation rate should be. The inflation rate is built into the mathematical definition of the currency, and it is self-limiting: the more work people put into mining bitcoins, the longer it takes to find new ones. This is not true of any existing currency, and is in fact an improvement on existing currencies.
Even if we don't all end up deciding to build an economy around bitcoin, this is an important achievement, something future currency designers ought to make note of.
What's to stop some economically significant entity (like a government) from starting a new Bitcoin-like pool of coins, with different parameters? This would allow them to control the overall supply of digital currency.
No, it doesn't offer any confidence that the bitcoin remains valuable. It could collapse entirely; falling prices could scare people out of bitcoins, lowering prices further, until bitcoins become absolutely worthless. This can't happen with commodity money because it has intrinsic value -- e.g. if you have an iron ore currency (backed by a ton of iron), it is valuable as long as iron is valuable. You can always sell your iron-dollar to someone who only wants the iron it is backed by. And a fiat currency will remain valuable (as long as it's not diluted by hyperinflation) because the state requires its residents to accept it as payment -- called "legal tender". But bitcoins? They are a bizarre sociological experiment -- a pyramid scheme, an enormous speculative bubble in something that has no intrinsic value. Or perhaps negative value, since bitcoins are costly to "create" -- they consume large amounts of CPU time.
What cryptography (reasonably) guarantees is that bitcoins aren't double-spent, aren't created from thin air. Not that they preserve value or are acceptable as payment.
As far as I understand it, the only value in any non-commodity-backed currency is a mutual agreement to use it as a medium of exchange for what are effectively IOUs. US dollars have no intrinsic worth, but I know that if you give me some, I can give them to someone else in exchange for something I want, and therefore I am willing to accept them in exchange for my labor. The only value in the dollar you give me is the fact that I can give it back to you later in exchange for something I actually want - but what makes it actual currency, and not just an IOU, is the fact that I can give it to anyone else in a dollar market in exchange for something I want.
Part of what makes this work is our mutual knowledge that it would be very hard for either of us to game the system. I can't just print up new dollars and neither can you; this means that the dollars you give me are very likely to have been given to you by someone else, which makes it very likely that I will be able to pass them on to someone else, which makes them useful and therefore valuable. If I could just print up my own dollars out of thin air, I wouldn't actually need to do any work to acquire them, which would make me less likely to accept them in trade for whatever it is you want from me; if everyone else is in the same position, then none of us have any reason to accept dollars from each other, and the currency is no longer useful as a means for exchanging debts. So the usefulness of any currency depends on the degree to which it is difficult to create new currency units.
This is what I mean by confidence that bitcoins are going to remain valuable: people can't just pull them out of thin air. It takes a lot of work to get a bitcoin. Right now the bitcoin system is new so lots of people are mining bitcoins, but once there are enough of them it will be easier to get a bitcoin by doing some other kind of work than by mining. This is all we need for bitcoins to become a useful medium of exchange for IOUs, should we decide to use it that way.
The fiat offers, at minimum, a requirement to do transactions in that currency to some minimal degree, as you are required by law to pay taxes (notice that taxes are applied to all exchanges, in theory at least, not just those conducted in the fiat currency) in that currency or go to jail.
That is the minimum it offers. It also offers the ability to transact with the government, who is the largest sole consumer in our economy, and a whole host of other things.
Did you think it's ok until someone used the word "lottery"? Would you feel better if it was called a "probabilistic" system instead? In reality the payouts are pretty predictable.
I don't gamble. This is not some religious thing, but simply because I like to have control over my fate, and gambling takes that control
Bu
Interesting attitude. But of course you don't have control over your fate. One day you're going to contract the disease which kills you, and you don't have any choice over when that happens or what it is. And between now and then, dozens of major events will happen in your life which you also didn't choose.
Maybe you like to have the illusion of control over your fate?
Personally I like to keep my life exposed to some randomness.
If you invest serious computing power into that, then there's not enough variance in it to be called gambling anymore. Betting on 6-6 in a single dice throw is gambling, betting on the percentage of 6-6's in a trillion dice throws is boring.
So if someone controls 51% of the BitCoin clients they can do anything they want?! Sounds like a flaw. Once lots of money is involved, it would make sense for people to start throwing botnets at this to hack it.
Well - the protocol is published and everyone can verify it for himself, the implementations are also open-source. This is not like someone is drawing a lot for you and you need to trust that he is doing that fairly - it is your computer that is doing that and it also checks that the others play by the rules.
Bitcoin should be rewritten to be some function of something "useful" like the SETI@Home project. Consider if you could have "solving sha256 problems" be simultaneously "computing the next largest prime" (arguably useful, at least, more so than what its currently doing) or "computing patterns in DNA".
The difficulty is that the problem must be of a predictable complexity. In other words, you must know ahead of time how long it takes a given amount of processing power to find the solution.
Would predictable statistical distribution of complexity not be enough? As in, you get a randomly assigned job, that takes on average a given amount of processing power, and is proven to not need more than some upper limit - or even, you time out after a given amount of processing power and request a new job.
And ideally have each unit small enough that mining will completing multiple jobs so difference in complexity starts averaging out.
You'd also need to be pretty certain that no-one's going to come up with a faster algorithm for solving the problem. I can't think of any real-world problems to which all these criteria would apply.
Someone please do the math on the energy lost in GPUs calculating hashes compared to the dollar value of the bitcoin rewards. The average miner might be better off just powering down their machines.
This math has already been done (see https://en.bitcoin.it/wiki/Power_Calc). Right now it's (theoretically) profitable to generate bitcoins; that would change if the exchange rate goes down.
That's making the assumption that the exchange rate goes down. If, (big if) the currency finds demand..the exchange rate goes up.
I saw it pointed out somewhere that bitcoins are meant to be rare, and that once the final bitcoin is generated, as there's meant to be a set ammount, there will be many orders of magnitudes less of them than dollars, euros, or the etc. So, even if the cost right now is high, having a hash of them could become worth a great deal.
That is of course speculative, but when there's going to be 1 bitcoin for every 600,000 of today's dollars...assuming the sum of bitcoins becomes worth a thousandth of that of dollars, each bitcoin will be worth 600 dollars...run a mining op for one coin a day for a year(365 coins), hold onto it for a while, average the price down to 300 dollars a coin...sell for a hundred grand...
So running a bitcoin mining opp, or even buying the coins at current market value, is a feasible long term high risk investment. I wouldn't suggest putting all your eggs in that basket, but I would suggest adding the purchase of the occasional bitcoin (or the cost of mining) to your list of high risk investments.
I've already started a bitcoin2 (actually it's called Bitcoin Prime).
The rules are as follows:
1. Every string of bits that is a bitcoin is also a Bitcoin Prime.
2. Currently I own all extant Bitcoins Prime, but I'm willing to sell them. Right now I am willing to sell them at a 50% discount relative to Bitcoins, so you can get twice the Bitcoins Prime for your US dollars.
You'd be an idiot to buy Bitcoin when you can buy Bitcoin Prime at half the price, right?
Never mind the energy; think of the capital expenditure. If I have US$1,000 lying around, I could either put it in a mutual fund, getting a certain number of dollars per month (with which I could buy bitcoins), or I could buy a kick-ass GPU card and become a miner, generating a certain number of bitcoins per month (which I could then sell in exchange for dollars). At current interest rates, which is a better investment?
Currently, mining is considerably more profitable. With $1000 you could probably put together a server that would generate 700 MHash/s, which currently equates to about $1000 in profit per month. You'd repay your initial investment quite quickly.
However, a mutual fund represents a consistent and safe income. I very much doubt that bitcoin mining will remain as profitable as it is now for very long.
Paypal's money market yeild is currently .06% per week.
At that rate, with $1000, you make .6 dollars a week, and buy maybe a bitcoin each month and a half...the discussion here is about earning a bitcoin a day mining, so I'm assuming you're down under a third of the mining return.
Assuming the cost of the electricitiy is 1.26 cents per kwh, and your machine is running at around 800 wats, constantly, that's 720 hours and 576 kwh, seven dollars a month...so the cost of running doesn't seem to be that high either.
Edited for bad math. May continue to do so. Please point out errors.
Long-term average stock market return is 5-8% depending on who you ask. So an index fund or balanced fund of funds would yield much better returns than bitcoin.
US Treasury Series I Savings bonds are currently paying 4.6% per year (their rate rises with inflation), income is tax exempt until redeemed, and can be redeemed at any time after 1 year.
I think you'd be better off going to the nearest bank with your $1,000 and buying an I bond. It's not a popular thing to do (only grandmas seem to like these things), but it's a pretty much slam dunk investment with zero risk.
Actually, even with a $200 ATI graphics card, you can earn about $500 revenue per month for about $100 in electrical costs.
That said, I suspect such a high level of profit won't last for long. It remains to be seen whether the current high value of bitcoins will last, and the more miners there are, the more profit has to be shared.
how'd you arive at $100 in electrical costs? that's rather higher than I calculated, and actually more than my home's entire monthly bill, with four computers being run sporadically, plus the central air nearly constantly, the washer, drier, water heater, refridgerator, tv, and other incedentals...
I was assuming a 700W power supply running 24 hours for 30 days, paying $0.20 per kwh. My cost estimates are admittedly a little on the high side, as I didn't want to be accused of overestimating the profit.
Perhaps true for the general case, but not true for people choosing to mine for Bitcoins. An single efficient modern GPU will use most of that 300W by itself when running full out.
True. I wasn't arguing for 700W with single card, but felt that 300W sustained might be a bit optimistic if one is aiming for $500 of Bitcoin revenue per month. The OP was offering a conservative estimate, and one could just as well point out that 20 cents/kWhr is also on the high side.
I think some of us are counting on the price to rise before we see significant return on investment. And for some others, electricity may be free, or at least not paid by themselves.
As for estimating the whole network's energy, here's a gross, probably exaggerated way to measure it. My 5870 gets 370 Mhash/s at 188 W. The current network rate is 2000 Ghash/s. 2000 / 0.37 * 188 = ~1 megawatt. Seems reasonable, especially since it's distributed worldwide.
I know I sound like a nutjob, but consider the carbon! A new currency built upon intensive coal hungry computing hardly seems like a long or even mid-term sustainable idea.
With the exchange rate atm being >7$/BTC, at the moment it is still profitable. As soon as it becomes unprofitable, I'll power down the machine. My miner costs me ~65EUR/month in electricity.
On the other hand, lots of other people will probably power down their miners, too, so that mining will become easier again. Then it could become feasible to mine again...
I am actually sorry about the waste of energy, but I decided that for once I'd join in the fun. I have never owned a car, so I hope that makes up for the waste to some extent.
a proper analysis might take into account the desirability of converting to estimate the number of miners which directly affects the cost of conversion. My guess is that as the number of bitcoins created starts to get rate limited, it will be right right below profitability over the long term because many miners can steal energy (e.g. a personal box at work) to warp the market.
I must admit I don't understand your first sentence. But it is true that many people can steal energy (student halls seem to be popular) and thus have an advantage in mining.
When I started my analysis was essentially "if I am lucky I can get back my investment and get a new gaming PC that pays for itself".
True, but that only means that my "people will quit and it will become profitable again" scenario might not happen. I think every miner assumes that the days of profitability are limited.
There I stood in front the infernal machine. Crumbled, soggy paper in hand. It fell to pieces before I even had the thought to preserve it somehow. Sixteen random characters, gone in the wash. My thesis, gone in the wash. A decade, gone in the wash.
I could try to rewrite it. But I wouldn't even know where to start. Of course all my research was encrypted too! I'm no idiot! This was work, this was hard work. I'm not even sure I could remember it all. "I wrote them down in my diary so I wouldn't have to remember."
I did the calculation. Brute forcing the password would take more than my lifetime with all the machines in the lab. Would Moore save me? Could I stand the wait? Later on, I stood on the roof of the math building (a quiet escape a few grad students shared) contemplating my next move. Looking out, late at night, it came to me. I couldn't do it, but the they could. Even just a fraction of the computers out there could get me my thesis back.
But how? I could create a distributed computing project, it was something I had worked on before. But there was no way to attract the required users. I needed something as least as big as the largest projects that already existed. I needed to give people something. A bounty isn't enough, the prime project has a great bounty, but I need more users. They all want money. But I had no money. So, I gave them the next best thing: the promise of money.
Please explain the intent of that graphic? Or, rather, what does where the electicity is have to do with voting power, aside from the congruence with cities?
No. There are already vastly superior anonymous e-voting schemes. The only usecase would be if you wanted to put some restriction on voting but still leave it open to everyone in the world, so it'd be kind of like Namecoin but paying for votes rather than DNS.
Related: here is how to calculate the expected number of days. Though I did not add in BitCoin specific parameters. But suppose you need to have zeros in the first k bits, I suppose the probability for each lottery ticket/number is (256-k)/256 (I assume that SHA256 has 256 bits, too lazy to look it up).
To me, at least, bitcoin isn't about a lottery and it isn't about winning. It's about laying the framework for the foundation of an open and secure currency. Who cares, in the long run, who won the most coins right now? I don't think many people have large stakes of bitcoins that they're hoping to get rich off of.
The mining cost should eventually be reflective of the value of the features of the cryptocurrency. In other words, if you want pseudoanonymous secure p2p transactions, buying bitcoins is like paying for that service. Who's providing the service? The miners, which are just the people keeping the network running so that others can use it.
Who cares, in the long run, who won the most coins right now?
In the long run, people with a lot of coins tend to get even more coins. I don’t see why I should participate in an “open and secure currency” that benefits people who let their hardware run wild doing mathematical operations which, outside the currency system itself, are pointless. At least gold has some aesthetic and industrial value.
But every new miner decreases the output of each mining rig. Posts about spending and using bitcoins benefit those who have bitcoins far more than "Free money" posts do.
Is there a point to these hashes? Does bitcoin just encourage people as a whole to make a trillion trillion hashes? I hope there'll never be a green bitcoin, it would be such a lie.
Can't someone adapt it to look for md5/sha clashes?
87 comments
[ 4.7 ms ] story [ 159 ms ] threadThere are economic problems with bitcoin as a currency (its a fiat currency, effectively, except without the fiat - it has scarcity with no intrinsic value), but the algorithm component seems solid to me.
For example, if an overwhelming majority of people with bitcoins(or USD) decides tomorrow to not ever use them again, then the remainder that have them are screwed. If the places that the currency can be uses cease to exist, the currency is worthless. There isn't any technical solution to this problem, the economy just ceases to exist. Obviously, this is an extreme case, but it's entirely possible that this can happen with any currency over an extended period of time if it doesn't catch on, or if investors lose confidence in the currency itself.
Part of what it means to have confidence in a currency is to be reasonably sure that its inflation will be limited to a rate that allows you to exchange your currency tokens before they are drained of value. Passing the hot potato, etc. With a fiat currency that confidence is based on nothing more than your beliefs about the central issuing agent's likely future actions.
What's new with bitcoin is that there is no central authority who can decide what the inflation rate should be. The inflation rate is built into the mathematical definition of the currency, and it is self-limiting: the more work people put into mining bitcoins, the longer it takes to find new ones. This is not true of any existing currency, and is in fact an improvement on existing currencies.
Even if we don't all end up deciding to build an economy around bitcoin, this is an important achievement, something future currency designers ought to make note of.
What cryptography (reasonably) guarantees is that bitcoins aren't double-spent, aren't created from thin air. Not that they preserve value or are acceptable as payment.
Part of what makes this work is our mutual knowledge that it would be very hard for either of us to game the system. I can't just print up new dollars and neither can you; this means that the dollars you give me are very likely to have been given to you by someone else, which makes it very likely that I will be able to pass them on to someone else, which makes them useful and therefore valuable. If I could just print up my own dollars out of thin air, I wouldn't actually need to do any work to acquire them, which would make me less likely to accept them in trade for whatever it is you want from me; if everyone else is in the same position, then none of us have any reason to accept dollars from each other, and the currency is no longer useful as a means for exchanging debts. So the usefulness of any currency depends on the degree to which it is difficult to create new currency units.
This is what I mean by confidence that bitcoins are going to remain valuable: people can't just pull them out of thin air. It takes a lot of work to get a bitcoin. Right now the bitcoin system is new so lots of people are mining bitcoins, but once there are enough of them it will be easier to get a bitcoin by doing some other kind of work than by mining. This is all we need for bitcoins to become a useful medium of exchange for IOUs, should we decide to use it that way.
That is the minimum it offers. It also offers the ability to transact with the government, who is the largest sole consumer in our economy, and a whole host of other things.
I don't gamble. This is not some religious thing, but simply because I like to have control over my fate, and gambling takes that control.
Maybe you like to have the illusion of control over your fate?
Personally I like to keep my life exposed to some randomness.
If you invest serious computing power into that, then there's not enough variance in it to be called gambling anymore. Betting on 6-6 in a single dice throw is gambling, betting on the percentage of 6-6's in a trillion dice throws is boring.
So if someone controls 51% of the BitCoin clients they can do anything they want?! Sounds like a flaw. Once lots of money is involved, it would make sense for people to start throwing botnets at this to hack it.
http://www.debtdeflation.com/blogs/2009/01/31/therovingcaval...
I saw it pointed out somewhere that bitcoins are meant to be rare, and that once the final bitcoin is generated, as there's meant to be a set ammount, there will be many orders of magnitudes less of them than dollars, euros, or the etc. So, even if the cost right now is high, having a hash of them could become worth a great deal.
That is of course speculative, but when there's going to be 1 bitcoin for every 600,000 of today's dollars...assuming the sum of bitcoins becomes worth a thousandth of that of dollars, each bitcoin will be worth 600 dollars...run a mining op for one coin a day for a year(365 coins), hold onto it for a while, average the price down to 300 dollars a coin...sell for a hundred grand...
So running a bitcoin mining opp, or even buying the coins at current market value, is a feasible long term high risk investment. I wouldn't suggest putting all your eggs in that basket, but I would suggest adding the purchase of the occasional bitcoin (or the cost of mining) to your list of high risk investments.
When bitcoins become rare why won't someone else start a new bitcoin2? I think there's an asymptotic max price that can't be crossed.
The rules are as follows:
1. Every string of bits that is a bitcoin is also a Bitcoin Prime.
2. Currently I own all extant Bitcoins Prime, but I'm willing to sell them. Right now I am willing to sell them at a 50% discount relative to Bitcoins, so you can get twice the Bitcoins Prime for your US dollars.
You'd be an idiot to buy Bitcoin when you can buy Bitcoin Prime at half the price, right?
However, a mutual fund represents a consistent and safe income. I very much doubt that bitcoin mining will remain as profitable as it is now for very long.
Assuming the cost of the electricitiy is 1.26 cents per kwh, and your machine is running at around 800 wats, constantly, that's 720 hours and 576 kwh, seven dollars a month...so the cost of running doesn't seem to be that high either.
Edited for bad math. May continue to do so. Please point out errors.
1) Since 1971 (40 years) is roughly 1300%, which is 19.63% annualized (1.1963^40 = 1300).
2) Since 1950 (60 years) is roughly 7680%, which is 16.08% annualized (1.1608^60 = 7680)
http://chart.finance.yahoo.com/z?s=%5eGSPC&t=my&q=l&...
I think you'd be better off going to the nearest bank with your $1,000 and buying an I bond. It's not a popular thing to do (only grandmas seem to like these things), but it's a pretty much slam dunk investment with zero risk.
That said, I suspect such a high level of profit won't last for long. It remains to be seen whether the current high value of bitcoins will last, and the more miners there are, the more profit has to be shared.
http://www.bitcoin.org/smf/index.php?topic=7692.0
As for estimating the whole network's energy, here's a gross, probably exaggerated way to measure it. My 5870 gets 370 Mhash/s at 188 W. The current network rate is 2000 Ghash/s. 2000 / 0.37 * 188 = ~1 megawatt. Seems reasonable, especially since it's distributed worldwide.
On the other hand, lots of other people will probably power down their miners, too, so that mining will become easier again. Then it could become feasible to mine again...
I am actually sorry about the waste of energy, but I decided that for once I'd join in the fun. I have never owned a car, so I hope that makes up for the waste to some extent.
When I started my analysis was essentially "if I am lucky I can get back my investment and get a new gaming PC that pays for itself".
There is a constant for M that is independent of P, M_power_leech.
I could try to rewrite it. But I wouldn't even know where to start. Of course all my research was encrypted too! I'm no idiot! This was work, this was hard work. I'm not even sure I could remember it all. "I wrote them down in my diary so I wouldn't have to remember."
I did the calculation. Brute forcing the password would take more than my lifetime with all the machines in the lab. Would Moore save me? Could I stand the wait? Later on, I stood on the roof of the math building (a quiet escape a few grad students shared) contemplating my next move. Looking out, late at night, it came to me. I couldn't do it, but the they could. Even just a fraction of the computers out there could get me my thesis back.
But how? I could create a distributed computing project, it was something I had worked on before. But there was no way to attract the required users. I needed something as least as big as the largest projects that already existed. I needed to give people something. A bounty isn't enough, the prime project has a great bounty, but I need more users. They all want money. But I had no money. So, I gave them the next best thing: the promise of money.
US power sources overlayed on 2008 Electoral Map. Too close to call.
http://i.imgur.com/d1Fyf.jpg
I was too lazy to fix Alaska and Hawaii.
Source: http://www.npr.org/templates/story/story.php?storyId=1109973...
http://blog.blinker.net/2011/02/08/expected-number-of-coin-t...
At the moment I generate about 5BTC/day with my dual-core ATI graphics card. Two months ago (when I started), it was twice that.
The mining cost should eventually be reflective of the value of the features of the cryptocurrency. In other words, if you want pseudoanonymous secure p2p transactions, buying bitcoins is like paying for that service. Who's providing the service? The miners, which are just the people keeping the network running so that others can use it.
In the long run, people with a lot of coins tend to get even more coins. I don’t see why I should participate in an “open and secure currency” that benefits people who let their hardware run wild doing mathematical operations which, outside the currency system itself, are pointless. At least gold has some aesthetic and industrial value.
Can't someone adapt it to look for md5/sha clashes?