A growing trend is to mine Bitcoin in Venezuela, where the electricity is hugely subsidized, mining can pay better than something like being a doctor, and you can get food sent in from Miami (since your income is in a widely accepted currency.)
In other words, the Bitcoin network consumes about 5-6 orders of magnitude more electricity than would be required to maintain a similarly-sized ledger using more traditional database techniques.
Yes, the feature set is a bit different, but it's difficult to see how Bitcoin can remain competitive in the long run. I guess it can help push down overheads elsewhere in the financial system, so at least that's a good thing. And it's always been a good educational experience.
The point of Bitcoin is to consume as much electricity, roughly, as a block is worth. Consuming this electricity, and using it to calculate a proof of the consumed electricity, means that the blockchain becomes immutable: no one can change the past, because it would require re-doing the proof-of-work.
So, out of this -- seemingly meaningless -- activity of burning off energy, arises the trustless property of the Bitcoin blockchain: with digital currencies that use proof-of-stake, or an SQL backend, every participant could -- theoretically -- have mutually agreed on assigning themselves 100x as many currency units, right before you enter and buy $1000 worth. There's no way to prove that this didn't happen, as it takes little effort to change history. This is unacceptable for a digital currency, hence the need for proof-of-work, and Bitcoin.
It's easy to change history with a plain SQL backend. But in any approach that stores the ledger in anything that boils down to a publicly distributed Merkle DAG, changing history or other manipulation like generating assets from nothing is impossible to do in secret.
The only potential issue is double-spending, or to be more precise, invalidating a transaction after the fact by pretending that a different, conflicting one, came first. This problem could be solved much more cheaply than in Bitcoin by using signatures by a trusted (e.g. elected) well-known council.
The only problem with that approach is how to bootstrap the value of such an alternative currency. Bitcoin somehow achieved that by appealing to a mixture of cypherpunk and libertarian ideology, and from there it developed to the current status which is an entirely circular origin to the value. I don't know what the alternatives for bootstrapping are. Fiat currencies can bootstrap via taxation, but that's not really an option here. Perhaps something like getting Valve to use such a system for their card trading :)
> The only problem with that approach is how to bootstrap the value of such an alternative currency
The problem of "bootstrapping the value" of a centralized, digital currency is that no one wants it, because of the reason I mentioned (too easy to change history). Working hard in order to earn an entry in some database, that can be deleted in a second, makes little sense for people.
Bitcoin had no bootstrapping problem because it can be used tustlessly (deleting past transactions is simply not possible), while your solution is just another type of credit. Why would I ever, voluntarily, start using your credit as money? It makes no sense from my perspective; hence the "bootstrapping problem".
In short: reducing the energy consumption of Bitcoin while reducing security is easy (use credit). It's keeping the trustless property while reducing energy usage that's the challenge.
I'm not sure what you mean by centralized, digital currency, but it's probably not what I was talking about, which involves a publicly distributed Merkle DAG.
Also, it's disingenious to pretend that Bitcoin had no bootstrapping problem. Bitcoin existed for a long time without significant value, and it seems stunts like spending what would now be millions of dollars for pizza were necessary pieces of solving the bootstrapping puzzle.
I'm referring to your solution to the double-spend problem, which is the problem Bitcoin solves in a trustless manner.
> This problem could be solved much more cheaply than in Bitcoin by using signatures by a trusted (e.g. elected) well-known council.
This is a centralized solution, where a group of -- albeit elected -- officials have complete control over the history of all balances. I'm arguing that no one is interested in such a system. No one wants democratic money, when it comes down it, is what I'm arguing. They want trustless money.
I'm also arguing that Bitcoin has monetary value in and of itself -- because it has use value -- and that its dollar price is a symptom of that, not the other way around. Even with zero dollars bidding on bitcoins, Bitcoin can be used for spam prevention -- out of which it was derived in the first place (Hashcash).
On the one hand, if each coin is worth 1/10th its cost, nobody will mint them. On the other hand, if each coin is worth 10x as much as it cost to make, people make their own and the price collapses.
That applies equally to gold coins as bitcoins.
I suspect bitcoin is, in its current form, incapable of replacing a major currency, because if it become a dominant currency then someone would beg, borrow, and steal just enough bitcoin to buy just enough compute power to get half of the bitcoin computer pool, from there using their new power to double-spend their way to getting even more computer power (while blocking other transactions that might try to out-bid them), and ultimately using that power to rewrite the consensus reality of the blockchain to say they now own as much as they want to say they own.
Even if the bitcoin wiki is correct to expect no rational miner will attempt this, there are plenty of irrational actors when enough money is dangled in front of them, even when it is only a mirage (citation: every lottery).
"if each coin is worth 1/10th its cost, nobody will mint them"
No. What will happen is as the miners with the highest mining costs drop out of the network, the difficulty target will decrease and decrease... until people stop dropping out of the network.
"people make their own and the price collapses"
No, again the difficulty target would adjust to regulate the rate of coin rewards.
Bitcoin is both the ledger and the printing/minting mechanism and the delivery mechanism (armored cars delivering new bills & coins to banks and picking up and destroying old ones).
Factor in those costs and suddenly bitcoin is bargain, energy-wise.
"it's difficult to see how Bitcoin can remain competitive"
Quite the contrary, the fact Bitcoin has been successful & growing for 8 years in spite of using so much energy is proof it is a non-problem. Energy per tx is rapidly decreasing (http://blog.zorinaq.com/bitcoin-mining-is-not-wasteful/) so it will only become more and more competitive from now on.
Let's hope it doesn't! Here's a Fermi problem to show how horrifying the prospect of Bitcoin succeeding is:
By this blog post's estimate, Bitcoin currently consumes about 500 MW of power. As others have estimated, Bitcoin processes about 3 transactions per second.
The world currently consumes about 2.38 TW of power.
If Bitcoin were a commonly-used currency -- if, for example, a billion people in the developed world each made an average of one Bitcoin transaction per day -- then Bitcoin would have to run 4,000 times faster than it currently does, consuming 2 TW of power, which is most of the power in the world.
You are incorrect. Bitcoin does not "have to" consume more energy to process more transactions. Energy and tx rate are uncorrelated. And the energy cost per transaction is falling. Again, see section 5 of http://blog.zorinaq.com/bitcoin-mining-is-not-wasteful/
Section 5 says "thanks to the transaction rate increasing faster than the network's energy consumption".
Is the transaction rate increasing? Two years ago people said it was 7 transactions per second. Now it's 3 per second. It may be that people were making unchecked exaggerations two years ago. But this doesn't look like any sort of increase. If the transaction rate is "quadrupling every 2 years"... why didn't it quadruple over 2 years?
An article entitled "Bitcoin Mining is Not Wasteful" on a Bitcoin blog kind of tells you what conclusion it's going to come to. For people who aren't personally invested in Bitcoin, of course Bitcoin mining is wasteful. It is literally made of wasted computation. The fact that it benefits an economy of Bitcoin speculators is of no interest.
It never was 7 tps. 7 is the theoretical maximum (assumes the tx had only 1 input.) In practice Bitcoin never achieves this because the average tx has multiple inputs.
"...tells you what conclusion it's going to come to"
You have to examine the presented facts, not the messenger. This blog is mine, and I think I present decent evidence that it's not wasteful. You have to look at the whole picture, ie. the Bitcoin economy lead to $1+ billion invested in 729+ companies, etc... Investing x megawatt in a system that produces investments and jobs is hardly a "waste".
> It never was 7 tps. 7 is the theoretical maximum
Wow, you mean Bitcoin fans were prone to exaggeration and unrealistic claims two years ago? I wonder if they still are.
Speaking of which, this graph you're showing me definitely has not quadrupled over the last two years. What went wrong?
> Investing x megawatt in a system that produces investments and jobs is hardly a "waste".
It is if the jobs never produce any actual value. Moving money around is not a good in itself. In fact a lot of Bitcoin companies have moved money in ways that are distinctly bad. Would you be celebrating MtGox as part of the Bitcoin economy if they still existed?
Of course whoever is currently holding the money is happy with it. Doesn't mean it's not a waste. People have profited from waste for as long as civilization has existed.
"Wow, you mean Bitcoin fans were prone to exaggeration and unrealistic claims two years ago?"
No one ever claimed it actually processed 7 tps.
"Speaking of which, this graph you're showing me definitely has not quadrupled over the last two years."
It quadrupled from 65k in Oct 2014 to 260k today. It's 2 years 5 months. Close enough. "2 years" was an approximation (perhaps I should edit my post to make it clear).
"It is if the jobs never produce any actual value"
Well I think they provide value. But it would be tedious & complex to argue about this on HN by reviewing the list of employees at 729+ companies...
Firstly MW = 10^6 and TW = 10^12 so you're off by 1,000. Secondly, your analysis doesn't factor indirect costs of centralised systems of currency (security features, printing, physical delivery, etc.).
In the first part of your reply, you indicate that you're bad at orders of magnitude, at least when defending Bitcoin.
In the second part of your reply, you do the same thing.
Let's be clear. (500 x 10^6) W x 4000 ~= 2 x 10^12 W. I will now factor in your fantasy that everyone stops using centralized currency: (500 x 10^6) W x (4000 - epsilon) ~= 2 x 10^12 W.
Fermi problems require being confident in how to multiply large numbers, and disregarding non-multiplicative factors that are much smaller.
if this happens, bitcoin would become valueless. The value of bitcoin comes from the fact that you can't just pump them out, the energy requirements required to solve the calculations required for mining are one of the limits to bitcoin production. Free energy like feasible solar or dyson spears would just be a license to print free digital currency making it worthless.
I think a rough estimate of the cost of 500MWh/year is $500 million per year. That doesn't include cooling, hardware, real-estate, staff and other operating expenses. I guess the majority of this cost is ultimately paid for by speculators who buy Bitcoins off of the miners.
$500 million per year is roughly what I get, too, but 500MWh/year is very different from 500MW, which is what I think you meant to write since the OP expresses its conclusion in MW.
If you accept that (1) there are economies of scale in Bitcoin mining and (2) miners are rational and will only mine when it is profitable, then doesn't this suggest that large amounts of mining power should become concentrated in the hands of just a few major miners? Doesn't that seriously undermine Bitcoin's validity?
"Each future halving of the Bitcoin block reward will narrow the range of profitable entry for
new Bitcoin miners considerably.
The stronger relative position of incumbents compared to that of new
entrants supports further consolidation of Bitcoin mining. Consequently, the distributed Bitcoin ledger, the blockchain, may become more vulnerable to attack, and its
trustless nature may ultimately even be destroyed."
" Once a predetermined number of coins have entered circulation, the incentive can transition entirely to transaction fees and be completely inflation free" - btc whitepaper
Transaction fees will take over mined bitcoins once there is non left to mine.
Okay but that does not solve the problem I am concerned about. That doesn't ensure a diverse mining pool. Already it doesn't make sense for me to mine Bitcoin (I don't have access to these ASICs). I think that economic forces will tend to concentrate mining power and that's a problem.
You are assuming the mining electricity comes predominately from fossil fuels which couldn't be further from reality. In reality the cheapest power is often in areas where hydroelectric is the dominate power source and miners choose their location based on availability of cheap power. Look at Washington State for example. It's a major producer of hydropower, has the cheapest electric rates in the US, and has a large number of Bitcoin mining facilities. Even in China it holds true that many mining facilities are in areas supplied by hydro electric power.
Indeed. I know of a medium-size 55nm Bitfury farm in Venezuela. But as others pointed out, most miners are not located in Venezuela. In fact Venezuelan authorities are cracking down on mining farms, and have shut this one down - https://news.bitcoin.com/venezuelan-authorities-weakening-bi...
> 100 gallons of gasoline also creates a ton of CO2 for context.
I thought I understood the Law of Conservation of Mass, but I guess I do not. If 100 gallons of gasoline has a mass of ~283.49 kg [0], how is a ton of carbon dioxide created from this?
Carbon atoms from the gasoline each combine with two oxygen atoms from the air to produce a carbon dioxide molecule, meaning the bulk of the mass of the resulting co2 actually comes from the atmosphere.
When an ~6 lb US gallon of no-ethanol gasoline is burned, it produces ~20 lbs of CO2 and ~8 lbs of H2O. The extra 22 lbs coming from oxygen in the air.
They make the assumption that unprofitable hardware has been retired. But this assumes that the person profiting from the mining is also paying for the electricity.
Good Econtalk on the subject: http://www.econtalk.org/archives/2017/02/jim_epstein_on.html