This argument comes up practically every day now with the same lack of comparison to existing financial systems. Bitcoin is an energy hog; so what? Lots of things require enormous amounts of energy. The question should be what value are we deriving from that energy usage, and is it greater than the status quo? How much energy is expended running bank servers? How much energy is expended mining gold and keeping reserves secured?
If bitcoin scaled to thousands of tps, then energy would be no brainer, and justified. Or if bitcoin lived up to its promises yet.
The argument I think we would probably agree on is some sort of revolution is happening, and there are consequences that will be forgotten once it ends. That i think people would agree on (setting aside if bitcoin is a revolution or a bubble).
What you are saying is "traffic accidents kill people", and then saying "so what - people die naturally every day!", and that's not OK.
Bitcoin is estimated to use about the same energy as the whole country of Denmark[1]. Denmark has over 790 million cashless transactions with "Dankort", a local debit card, per year alone[2] - excluding other payment methods. Bitcoin can handle about 2,000 transactions per block[3]. 2,000 transactions times 6 blocks per hour times 24 hours in a day times 365 days in a year equals 105,120,000 transactions per year.
Congratulations, you just used the energy of the whole country of Denmark with _everything_ in it (housholds, heavy industry, etc.) to process less than a 1/7th of Denmark's cash-less transactions with a _specific_ (albeit widely used) debit card.
You see, your argument is simply invalid. "Traditional" banking is more energy efficient by multiple orders of magnitude.
There are some very huge differences between Bitcoin and Denmark’s payment system. It’s perfectly fine for you to dislike or scoff at those differences, but some people find Bitcoin’s differences extremely appealing, and it’s disingenuous to say that the payment systems are accomplishing the same thing just with different energy costs.
The title of the article is yet another stab at the bitcoin energy usage calculation but the content is actually much more interesting. There's an assertion, that I can't validate, that there's a point at which mining becomes unsustainable because the energy cost becomes too expensive and thus the network crashes. Anyone have a rebuttal for that?
No. (Well, in the short term as miners leave, the block time increases, which increases the block pressure and increases fees, but the block time re-normalizes itself regularly as the mining difficulty adjusts.)
The only thing that the number of people mining directly affects is the cost for an attacker to perform a 51% attack on the network.
Yes, it is an energy hog. There are a lot of industries that consume more energy than bitcoin mining.
The reason none of this matters is because the market says it doesn't matter. The market is finding enough value from Bitcoin to keep running the network. Going forward, I think energy consumption won't matter because we are now producing more power than we can consume. For example, Germany had to pay citizens to consume power because there was a glut. Another example recently was in California where the state had to pay Arizona to take their excess capacity.
The same is happening in China. China is heavily investing in alternate forms of energy and when you boil it down, it is still cheap to mine a 14,000+ USD bitcoin. In other words, the market says it's ok.
Sure, I agree with that. It's not clear the comment I'm replying to does so. If we want to view the market as a phenomenon like soil erosion that's fine. I got the impression that the parent comment was implying markets as a provably optimal solver for resource allocation problems.
Sure, but when someone says that, what does it mean really? Market is the best process we have for determining relative value, just like science is the best process we have for determining truth.
But let's not take this for granted and let's try and approach this practically, for this specific subject.
If I said "I think the market is wrong and having a decentralised consensus system that provides strong cryptographic guarantees is not worth X amount of electricity, while maintaing multipple centralised banking systems is worth Y amount of electricity"
then there are some hard follow-up questions to deal with:
Is it not worth any amount of electricity then?
Why is it worth only that amount (or none)?
Who is to say what amount of electricity it is worth? So far we know that markets are the authotity of how much anything is worth. Why should we trust some other authority? And if we accept that some authority can decide how much electricity one application deserves, then why should it do it selectively? Why not say how much every application should be using?
And if we somehow found answers to the above, then we have to ask:
How would that be enforced?
Would we want to live in a society with this level of control where enforcing such things is possible?
Without extending to address these tough problematics, any opinion about the market being wrong carries no weight -the simple answer would just be
"if the market can be wrong, so can anything or anyone else we know of"
'Market is the best process we have for determining relative value, just like science is the best process we have for determining truth.'
I cant upvote this enough. People have been saying the price is what the market is willing to pay. This never stuck is as an answer, but both word relative AND value perfectly describe pricing. Thank You
None of the things my grandparent mentioned are externalities.
The cost of energy consumed is directly borne by the people mining the bitcoins. The costs of the power gluts should be being directly borne by the customers (I'm pretty sure it is in Germany, I don't know about California).
The energy market is distorted by government interference, sure, but the costs are clear and (unless the regulators are doing something really stupid) being borne by parties with skin in the game.
I think the best approach to internalizing those externalities are to build it into the cost of electricity. If electricity is too cheap, then it’s too cheap for everything, not just for Bitcoin mining.
> For example, Germany had to pay citizens to consume power because there was a glut
That would have been great. The opposite is the case: industrial users get paid and regular consumers pay those payouts with their monthly bill. Those payouts are a cost to the utilities
The market does not say it is "ok". It merely describes, in terms of USD, the amount of electricity and other resources utilized by Bitcoin. The market is amoral.
I didn't read the GP ascribing morality into that statement. They're simply saying that the energy consumption is a market effect -- if it wasn't in the miner's self-interest to mine, they wouldn't.
I read it the same as you, except I hate that line of reasoning to hand-wave things away.
Let us assume a perfectly rational market... should preface all of these statements. That way it triggers us appropriately like "Assume an infinite plane" does.
It matters quite a bit because of the halving. The security of bitcoin essentially relies on the idea that the electicty_cost_per_block/2 * num_conf_needed > tx_value_by_individual_actor.
e.g. if a block costs ~70k USD of electricity and you accept a tx for more than 210k USD after 6 conf then if the sender is a purely self interested actor in a perfect market their rational decision would be to spend a bit more than 210k USD in electricity to mine 7 blocks.
Obviously this is a completely idealized market where you can purchase mining power to fork at will.
But the end result is that the more electricity burned the larger tx you can trust with a lower amount of confirmations. In ~7 years we'll see block rewards drop to ~3 BTC. By that point if we don't have tx fees that compensate then we'll see a large drop in hash rate. If the size of the blocks doesn't increase, the size of a tx decrease or something else we will either see the security drop quite a bit or the cost of a on the chain tx increase dramatically.
> e.g. if a block costs ~70k USD of electricity and you accept a tx for more than 210k USD after 6 conf then if the sender is a purely self interested actor in a perfect market their rational decision would be to spend a bit more than 210k USD in electricity to mine 7 blocks.
The math doesn't work that way.
If you take all the bitcoin mining operations in the whole world, China and every other country, put them together and somehow duplicated all that under your own control, so that you own as much as everyone else put together, you own 50% of the global hashrate... that only gives you a 50% chance that you will mine the next block, let alone 7 in a row.
The chances of mining 7 blocks in a row, even if you own as much as everyone else in the world put together, is less than 1%.
And it would cost billions of dollars worth of ASICs to get that 1% chance.
You have this wrong. If you have 51% of the mining power, you do indeed have a negligible chance of winning seven consecutive blocks in a row if you mine the normal way. But, if you just mine seven sequential blocks without publishing the first six until you've mined all seven, you have a rather good chance of getting all seven blocks before the rest of the network mines seven blocks.
The idea that it would cost billions of dollars worth of ASICs to get this chance relies on the fact that Bitcoin is in an absurd boom right now, such that only new ASICs are competitive and there isn't a pile of old used ASICs available. If Bitcoin ever stagnates, this will rapidly stop being true.
You're also assuming that the ability to rewrite an hour of Bitcoin history isn't worth a couple billion dollars to anyone. That's an interesting assumption. (If I were, say, Israel, one of my enemies started using Bitcoin to buy weapons, and I thought that the ability to rewrite Bitcoin history an hour at a time could bankrupt them, I would be thrilled at the opportunity to spend a billion dollars to do so.)
I'm frankly rather surprised that Bitcoin, and proof-of-work blockchains in general, are considered secure enough for serious high-value work.
Nope, if you mine seven blocks in the time that it takes the rest of the network to mine six and you publish your longer chain, then the blockchain reorganizes and you win.
owning 50% + eplison of the hash rate gives you the expectation of being able to undo an unlimited amount of history in theory.
In practice you'd need to show the world the blocks before they got to old or they'd be dropped due to an old timestamp.
Buying up the ASICs would be the real cost, It seems like the chain is largely safe because of that. If the price drops enough that lots of mines liquidate it'd be good to keep an eye on where they go. If 50% of the hash power is dark large txs should rely on a whole lot of confirmations if at all.
I'm with you about half way. Sure, if we produce more power than we consume and we aren't using carbon-emitting power sources then there's no reason to dial this back so long as miners are willing to pay for the power and infrastructure to keep it going.
The problem of course is that China is still getting most of its power from fossil fuels (the rosy numbers they often promote usually end with "by 2050"). Given that the environmental effects of bitcoin's fossil fuel usage may not be borne by the people buying/mining bitcoin, it doesn't make sense to expect the market to price it in.
Edit: I mention China because that's where a lot of bitcoin is mined due to cheap electricity.
> For example, Germany had to pay citizens to consume power because there was a glut. Another example recently was in California where the state had to pay Arizona to take their excess capacity.
One off events that happen because of an alignment of several external factors are not good markers to plan the future from. Further, you can't assume the Bitcoin is the only sink for excess energy and that other markets/systems aren't in a better position to take advantage of it.
The thing about value is that you value things differently than other people. Thank goodness you, me, or any one person doesn’t get to decide to ban or penalize anything they don’t value.
I just saw this "you have to articles" thing for the first time. Is medium putting up a paywall or am I being A/B tested? Or maybe this was announced before and I'm out of the loop.
I've been a holder and believer since 2012 but this is one thing where I've always disagreed with the community--not so much the ecological aspect but the value argument based on production. Bitcoin's overwhelmingly libertarian user base had no time for this kind of discussion because it was too close to the labor theory of value.[1] But if you work it out you can actually get a solid argument for a current "intrinsic" value, taking into account expected future returns, etc. We're starting to see this with the markets for mining contracts where the unit is (mega/giga) hashes per second. Seems to be homologous to a forward contract.
But in any case the production aspect seems to have been totally ignored when it comes to valuation whereas with every other commodity--oil, corn, cotton, copper, etc.--it means everything.
>The next halving could cause a transient effect in the network where the number of machines mining on the network drops dramatically. This may have the effect of miners selling off their Bitcoin and the value falling.
This is not how markets work. Anything knowable about the future is reflected in the current price.
Only under efficient market hypothesis (EMH), which is not necessarily the true state of the market. Under EMH, 2008 would never have happened, because all the (unforeseen) risk would have been priced in as 'knowable' - the risk of default was mathematically provable, even if no one ever took the time to do the math.
At the point where all the bitcoins are mined, what is the incentive for 'miners' to keep processing transactions?
1) Parties who wish to transact must provide that incentive via paying transaction fees
2) The fees will be in proportion to the hash rate which is a function of the computing power/energy allocated to hashing which is related to the number of 'miners'
3) So the transaction fees will be in proportion to the cost of energy.
4) BUT ALSO anyone who owns bitcoin is incentivized to engage in 'mining' because as the number of 'miners' approaches zero, the blockchain becomes susceptible to attack (transactions are 'forged' and bitcoins are 'stolen') by a sufficiently powerful malicious party (51% attack)
5) So bitcoin's value will be a function of the interplay between the desire to transact with bitcoin and the desire to retain ownership of bitcoins
First if you are going to say something is correlated it'd be nice to provide an R^2 or something to quantify the correlation. Then given that price of bitcoin and cost of electricity goes back much further than what is listed. Has the correlation been consistent? (given the fact that bitcoin has been growing exponentially I would be surprised)
Second if you are trying to predict what will happen to the price & hashrate at the next halving it might help to look to previous examples. This has happened twice to bitcoin alone and many more times on various different coins. This will also affect other SHA256 PoW coins such as bitcoin cash, which is ~9000 blocks ahead. Interestingly it's halving will happen ~9 weeks earlier, not much lead time, but it should give an indication of what will happen to BTC.
Lastly:
> To compute cost of creating one Bitcoin, I established a correlation between the number of TH/day and kWh/day as value conv in TH/kWh. To compute the value use: 606024TH_sconv*energy_cost/1600
Huh?
Where does the 1600 value come from? There are currently 1800 new bitcoins minted every day, is that what it is supposed to be? If that's the case that is the cost to get one of the newly minted bitcoins.
Furthermore it's an extremely common misconception that if you spend X electricity you produce Y bitcoins. Saying that there is a specific electricity cost for creating a bitcoin could easily reinforce this idea. It looks like the author understands how it actually works, but the wording used seems like it could easily reinforce that misconception.
The latest ASIC miners run at about 100W/TH. That means the entire hashing capacity of Bitcoin (15 EH/s) on the most efficient hardware would draw 100,000 kW.
That’s about 80,000 US households of power. Or to put it another way, ~$21 million per month in electricity.
Obviously average mining power efficiency is currently much lower, so this isn’t much more than a thought exercise.
But the point is there are a lot of uses for electricity. The only real problem is if the price of electricity doesn’t actually cover the cost of the externalities of generating it.
In many places, you are paying not only for externalities but also infrastructure upgrades throughout the grid as well as subsidizing low income use, so in that sense Bitcoin is funding these positive programs and investments.
If the electricity is being sold below real market cost, don’t blame Bitcoin (or incandescent bulbs) blame the fucking electrical pricing regulations.
> The only real problem is if the price of electricity doesn’t actually cover the cost of the externalities of generating it.
We don't even know all the externalities of generating it, let alone adding all of them to the price.
E.g, nuclear waste and depleting fossil fuels are two well-known externalities which we currently consciously choose to ignore. Climate change is an externality which until about half of a century ago wasn't known at all.
Sure, we don’t know all the externalities, but some of those externalities could be positive. The fact that we don’t know doesn’t mean the true public cost of electricity is even higher than the market price.
Yes, you can say that about everything. "Maybe polluting the air and driving a mass extinction event has some surprising positive consequences we just don't know yet. We should go on and see if we discover them!"
There are plenty of hints that we don't correctly account even for the known negative externalities. There are also incidents of new negative externalities showing up we hadn't known about before. I don't know about hints for positive externalities in the same order of magnitude.
While I find your post informative, I don't agree with your last statement. Bitcoin is designed around the idea of burning energy (in the form of energy-expensive computation) as "proof of work." This design decision is certainly a valid point of criticism of the system.
Great point! My understanding is that proof of work is the only proven way to resist Sybil attacks in a fully distributed manner to-date.
When someone invents a better mechanism they will certainly have a huge economic advantage if you can join and support the miner network without consuming so many real resources.
Decentralized Sybil resistance at this point requires a hard cost to limit spam. In a fully decentralized system, this seems like a natural law that this must be true. I have ideas on how to solve Sybil without resource exhaustion, but they are just at the initial prototype phase.
If the critique is that Bitcoin specifically isn’t worth the “cost” of all the electricity, it’s a moral question which is easily solved by better pricing electricity. If the critique is Bitcoin is wasteful, then point to a more efficient algorithm or it’s tilting at windmills.
There are algorithms which aren’t as resource intensive but the market judged the trade-offs in centralization to not be worth the benefits. We’ve know how to store a centralized ledger for a long time, so it’s totally reasonable that steps towards centralization for efficiency sake are a non-starter.
I don't think it's unreasonable to argue centralization is better because it in principle allows creation of currency without the requisite resource requirements.
This is a reasonable argument to have. Especially because the economies of scale from mining power may lead to something a little like centralization anyways.
It is fully reasonable to argue that the supposed benefits of decentralization via a Bitcoin-like system do not outweigh the benefits of centralization in the real world. As far as the "market deciding," well currently the market values centralized fiat currency far more than decentralized cryptocurrency, so I don't think that's a very strong argument. (Plus, "the market can stay irrational longer than you can stay solvent.")
With PoW + self-adjusting difficulty, bitcoin is quite intentionally designed to encourage ever-higher energy consumption to "secure the blockchain". Yet HN is trying to "well actually"-away the fact that bitcoin is an energy hog.
> bitcoin is quite intentionally designed to encourage ever-higher energy consumption to "secure the blockchain"
It's designed to have have an average of 10 minute blocks using the feedback of cumulative proof of work over a given amount of blocks compared to a difficulty parameter.
The only reason it uses more energy is because people believe there is money to be made in mining. They will be wrong sometimes, and right sometimes, and equilibrium will be reached intermittently. This is called a market.
And yet in result, this market is justifying that the exact same task (verifying a block every 10 min) needs orders of magnitude more energy today than it did in the past, even though the result is exactly the same. This is practically an energy hog.
I'd argue that bitcoin is also an energy hog in theory:
Why do people believe there is money to be made in mining? Because bitcoin's core mechanisms ensure that money is to be made, as long as bitcoin is worth enough (which it currently is).
So the bitcoin protocol does three things:
1) Encourage people to mine by offering block rewards and tx fees.
2) Ensure that new miners don't actually add any useful capacity to the network.
3) Require that old miners do at least as much work as new miners if they want to keep their income.
If you take those together, the bitcoin protocol encourages a race to spend more energy to archieve the same thing.
> They will be wrong sometimes, and right sometimes, and equilibrium will be reached intermittently.
There is an equilibrium of energy costs in bitcoin, obviously. However no one knows how the bitcoin/USD price will develop, so I don't see how that implies any cap for energy costa in USD. And even that is just an aporoximation for the actual energy costs in raw materials. One that can be very skewed as the posts further down in the thread explain.
Let people be wasters, nothing wrong with that. If Bitcoin experiment succeeds it's going to be better than the Internet. I hear no one complaining that Internet servers waste magnitudes more energy than Bitcoin, serving porn, FB, flash games and all other wasters of life.
This is an interesting theoretical implication of cryptocurrencies' decrentralised mining process. The welfare of monopoly vs. competitive markets are the opposite to traditional micoroeconomics, because a currency's social value does not increase with mining effort.
Any central bank can use their monopoly over the creation of currency to increase reserves whilst incurring a cost less than those reserves. For example, suppose the US economy grows by 50% over 30 years, but that the country's payment infrastructure doesn't change otherwise, and so the typical American wishes to have a higher real value of cash in their wallet, to buy more luxurious office lunches, or spend the cash in other circumstances.
In a situation where everyone uses a traditional fiat currency, minted by a central bank, the central banks monopoly over the production of these additional currency reserves will allow them to spend significantly less than the value of the currency in making it. Today, a $100 bill costs 12.5 cents to make, plus some additional transport and monitoring costs [1].
However, with the decentralised nature of a crypo-currency, a competitive market dictates that in order for it to be created, miners must spend the entire value of the cryptocurrency in energy costs, otherwise they would be forgoing risk-free profits. Suppose the economy uses cryptocurrency. If the market allows for miners to spend only $90 of energy, hardware and labour to mine $100 of currency, then a less efficient entrant can come in with costs between $90-$100 and make a positive expected profit. This means that even if there are not technical difficulties preventing a crypocurrency from replacing a traditional fiat currency, it's adoption reduces the social surplus generated from the minting of new currency.
OT: A friend of mine is trying to enter into crypto-market with slightly different idea/project but related to blockchain. I'm not a crypto expert; if anyone feels like helping with reviewing his basic idea, please feel free to reach out via email in my profile. This could be somewhat paid small (few-times) gig. /OT
Preface: I'm a 2.25MW bitcoin miner myself (see https://toom.im).
This author is getting the causation backwards:
> Is there a correlation between power consumption and the cost of bitcoin.
Yes, of course there is a correlation, but he's getting the causation entirely backwards.
The profitability of mining is directly proportional to the price of bitcoin, because you're paid with 12.5 Bitcoins per block. But your costs are fixed in USD -- you pay for electricity costs, per kilowatt. Therefore as the Bitcoin price goes up, it's profitable to run that many more miners, to produce that many more hashes, until you fill up the profit with electricity costs again.
So the price of Bitcoin determines the hashrate, not the other way around, like he's presuming here:
> If this correlation is strong enough, it means the Bitcoin will always be bounded by the cost and availability of electricity.
This guy also assumes that miners are paying the average electricy price in the US and China. Miners pay nothing like that. California electricity rates are around 15¢/kW. Central Washington is 2-4¢. That's 6-7x cheaper. The same is true for China.
74 comments
[ 3.0 ms ] story [ 114 ms ] threadA neutral headline would just talk about bitcoin energy consumption or whatever.
A better question is "what value does Bitcoin provide for the energy it uses compared to similar alternatives (Ethereum, Ripple...)?"
The argument I think we would probably agree on is some sort of revolution is happening, and there are consequences that will be forgotten once it ends. That i think people would agree on (setting aside if bitcoin is a revolution or a bubble).
What you are saying is "traffic accidents kill people", and then saying "so what - people die naturally every day!", and that's not OK.
Congratulations, you just used the energy of the whole country of Denmark with _everything_ in it (housholds, heavy industry, etc.) to process less than a 1/7th of Denmark's cash-less transactions with a _specific_ (albeit widely used) debit card.
You see, your argument is simply invalid. "Traditional" banking is more energy efficient by multiple orders of magnitude.
1: https://arstechnica.com/tech-policy/2017/12/bitcoins-insane-...
2: https://www.nationalbanken.dk/en/publications/Documents/2011... (page 125)
3: https://blockchain.info/en/charts/n-transactions-per-block
'Dankort' isn't global
Can I divide a Dankort into smaller Dankort units?
Can I send my friend a Dankort? Can she send a Dankort to me?
Citibank uses more electricity
When people stop mining, competition for mining falls, making it easier to mine.
If it's easier to mine, mining becomes cheaper, so people start mining.
GOTO 1
The only thing that the number of people mining directly affects is the cost for an attacker to perform a 51% attack on the network.
The reason none of this matters is because the market says it doesn't matter. The market is finding enough value from Bitcoin to keep running the network. Going forward, I think energy consumption won't matter because we are now producing more power than we can consume. For example, Germany had to pay citizens to consume power because there was a glut. Another example recently was in California where the state had to pay Arizona to take their excess capacity.
The same is happening in China. China is heavily investing in alternate forms of energy and when you boil it down, it is still cheap to mine a 14,000+ USD bitcoin. In other words, the market says it's ok.
Do you think the precious metal markets are wrong?
Markets aren't right or wrong. People decide what's right and wrong, and apply the relevant constraints via regulation.
But let's not take this for granted and let's try and approach this practically, for this specific subject.
If I said "I think the market is wrong and having a decentralised consensus system that provides strong cryptographic guarantees is not worth X amount of electricity, while maintaing multipple centralised banking systems is worth Y amount of electricity"
then there are some hard follow-up questions to deal with:
Is it not worth any amount of electricity then?
Why is it worth only that amount (or none)?
Who is to say what amount of electricity it is worth? So far we know that markets are the authotity of how much anything is worth. Why should we trust some other authority? And if we accept that some authority can decide how much electricity one application deserves, then why should it do it selectively? Why not say how much every application should be using?
And if we somehow found answers to the above, then we have to ask:
How would that be enforced?
Would we want to live in a society with this level of control where enforcing such things is possible?
Without extending to address these tough problematics, any opinion about the market being wrong carries no weight -the simple answer would just be
"if the market can be wrong, so can anything or anyone else we know of"
'Market is the best process we have for determining relative value, just like science is the best process we have for determining truth.'
I cant upvote this enough. People have been saying the price is what the market is willing to pay. This never stuck is as an answer, but both word relative AND value perfectly describe pricing. Thank You
https://www.newyorker.com/cartoon/a16995
The cost of energy consumed is directly borne by the people mining the bitcoins. The costs of the power gluts should be being directly borne by the customers (I'm pretty sure it is in Germany, I don't know about California).
The energy market is distorted by government interference, sure, but the costs are clear and (unless the regulators are doing something really stupid) being borne by parties with skin in the game.
That would have been great. The opposite is the case: industrial users get paid and regular consumers pay those payouts with their monthly bill. Those payouts are a cost to the utilities
Let us assume a perfectly rational market... should preface all of these statements. That way it triggers us appropriately like "Assume an infinite plane" does.
e.g. if a block costs ~70k USD of electricity and you accept a tx for more than 210k USD after 6 conf then if the sender is a purely self interested actor in a perfect market their rational decision would be to spend a bit more than 210k USD in electricity to mine 7 blocks.
Obviously this is a completely idealized market where you can purchase mining power to fork at will.
But the end result is that the more electricity burned the larger tx you can trust with a lower amount of confirmations. In ~7 years we'll see block rewards drop to ~3 BTC. By that point if we don't have tx fees that compensate then we'll see a large drop in hash rate. If the size of the blocks doesn't increase, the size of a tx decrease or something else we will either see the security drop quite a bit or the cost of a on the chain tx increase dramatically.
The math doesn't work that way.
If you take all the bitcoin mining operations in the whole world, China and every other country, put them together and somehow duplicated all that under your own control, so that you own as much as everyone else put together, you own 50% of the global hashrate... that only gives you a 50% chance that you will mine the next block, let alone 7 in a row.
The chances of mining 7 blocks in a row, even if you own as much as everyone else in the world put together, is less than 1%.
And it would cost billions of dollars worth of ASICs to get that 1% chance.
The idea that it would cost billions of dollars worth of ASICs to get this chance relies on the fact that Bitcoin is in an absurd boom right now, such that only new ASICs are competitive and there isn't a pile of old used ASICs available. If Bitcoin ever stagnates, this will rapidly stop being true.
You're also assuming that the ability to rewrite an hour of Bitcoin history isn't worth a couple billion dollars to anyone. That's an interesting assumption. (If I were, say, Israel, one of my enemies started using Bitcoin to buy weapons, and I thought that the ability to rewrite Bitcoin history an hour at a time could bankrupt them, I would be thrilled at the opportunity to spend a billion dollars to do so.)
I'm frankly rather surprised that Bitcoin, and proof-of-work blockchains in general, are considered secure enough for serious high-value work.
Then you end up with an orphaned chain which will have effect at all.
In practice you'd need to show the world the blocks before they got to old or they'd be dropped due to an old timestamp.
Buying up the ASICs would be the real cost, It seems like the chain is largely safe because of that. If the price drops enough that lots of mines liquidate it'd be good to keep an eye on where they go. If 50% of the hash power is dark large txs should rely on a whole lot of confirmations if at all.
The problem of course is that China is still getting most of its power from fossil fuels (the rosy numbers they often promote usually end with "by 2050"). Given that the environmental effects of bitcoin's fossil fuel usage may not be borne by the people buying/mining bitcoin, it doesn't make sense to expect the market to price it in.
Edit: I mention China because that's where a lot of bitcoin is mined due to cheap electricity.
One off events that happen because of an alignment of several external factors are not good markers to plan the future from. Further, you can't assume the Bitcoin is the only sink for excess energy and that other markets/systems aren't in a better position to take advantage of it.
So what? That doesn't make its energy use any less unethical. Especially since the only "value" that Bitcoin provides is as a tool for speculation.
But in any case the production aspect seems to have been totally ignored when it comes to valuation whereas with every other commodity--oil, corn, cotton, copper, etc.--it means everything.
[1] https://en.bitcoin.it/wiki/Help:FAQ#Where_does_the_value_of_...
This is not how markets work. Anything knowable about the future is reflected in the current price.
1) Parties who wish to transact must provide that incentive via paying transaction fees
2) The fees will be in proportion to the hash rate which is a function of the computing power/energy allocated to hashing which is related to the number of 'miners'
3) So the transaction fees will be in proportion to the cost of energy.
4) BUT ALSO anyone who owns bitcoin is incentivized to engage in 'mining' because as the number of 'miners' approaches zero, the blockchain becomes susceptible to attack (transactions are 'forged' and bitcoins are 'stolen') by a sufficiently powerful malicious party (51% attack)
5) So bitcoin's value will be a function of the interplay between the desire to transact with bitcoin and the desire to retain ownership of bitcoins
Thoughts?
First if you are going to say something is correlated it'd be nice to provide an R^2 or something to quantify the correlation. Then given that price of bitcoin and cost of electricity goes back much further than what is listed. Has the correlation been consistent? (given the fact that bitcoin has been growing exponentially I would be surprised)
Second if you are trying to predict what will happen to the price & hashrate at the next halving it might help to look to previous examples. This has happened twice to bitcoin alone and many more times on various different coins. This will also affect other SHA256 PoW coins such as bitcoin cash, which is ~9000 blocks ahead. Interestingly it's halving will happen ~9 weeks earlier, not much lead time, but it should give an indication of what will happen to BTC.
Lastly:
> To compute cost of creating one Bitcoin, I established a correlation between the number of TH/day and kWh/day as value conv in TH/kWh. To compute the value use: 606024TH_sconv*energy_cost/1600
Huh?
Where does the 1600 value come from? There are currently 1800 new bitcoins minted every day, is that what it is supposed to be? If that's the case that is the cost to get one of the newly minted bitcoins.
Furthermore it's an extremely common misconception that if you spend X electricity you produce Y bitcoins. Saying that there is a specific electricity cost for creating a bitcoin could easily reinforce this idea. It looks like the author understands how it actually works, but the wording used seems like it could easily reinforce that misconception.
That’s about 80,000 US households of power. Or to put it another way, ~$21 million per month in electricity.
Obviously average mining power efficiency is currently much lower, so this isn’t much more than a thought exercise.
But the point is there are a lot of uses for electricity. The only real problem is if the price of electricity doesn’t actually cover the cost of the externalities of generating it.
In many places, you are paying not only for externalities but also infrastructure upgrades throughout the grid as well as subsidizing low income use, so in that sense Bitcoin is funding these positive programs and investments.
If the electricity is being sold below real market cost, don’t blame Bitcoin (or incandescent bulbs) blame the fucking electrical pricing regulations.
We don't even know all the externalities of generating it, let alone adding all of them to the price.
E.g, nuclear waste and depleting fossil fuels are two well-known externalities which we currently consciously choose to ignore. Climate change is an externality which until about half of a century ago wasn't known at all.
There are plenty of hints that we don't correctly account even for the known negative externalities. There are also incidents of new negative externalities showing up we hadn't known about before. I don't know about hints for positive externalities in the same order of magnitude.
When someone invents a better mechanism they will certainly have a huge economic advantage if you can join and support the miner network without consuming so many real resources.
Decentralized Sybil resistance at this point requires a hard cost to limit spam. In a fully decentralized system, this seems like a natural law that this must be true. I have ideas on how to solve Sybil without resource exhaustion, but they are just at the initial prototype phase.
If the critique is that Bitcoin specifically isn’t worth the “cost” of all the electricity, it’s a moral question which is easily solved by better pricing electricity. If the critique is Bitcoin is wasteful, then point to a more efficient algorithm or it’s tilting at windmills.
There are algorithms which aren’t as resource intensive but the market judged the trade-offs in centralization to not be worth the benefits. We’ve know how to store a centralized ledger for a long time, so it’s totally reasonable that steps towards centralization for efficiency sake are a non-starter.
This is a reasonable argument to have. Especially because the economies of scale from mining power may lead to something a little like centralization anyways.
It is fully reasonable to argue that the supposed benefits of decentralization via a Bitcoin-like system do not outweigh the benefits of centralization in the real world. As far as the "market deciding," well currently the market values centralized fiat currency far more than decentralized cryptocurrency, so I don't think that's a very strong argument. (Plus, "the market can stay irrational longer than you can stay solvent.")
With PoW + self-adjusting difficulty, bitcoin is quite intentionally designed to encourage ever-higher energy consumption to "secure the blockchain". Yet HN is trying to "well actually"-away the fact that bitcoin is an energy hog.
It's designed to have have an average of 10 minute blocks using the feedback of cumulative proof of work over a given amount of blocks compared to a difficulty parameter.
The only reason it uses more energy is because people believe there is money to be made in mining. They will be wrong sometimes, and right sometimes, and equilibrium will be reached intermittently. This is called a market.
I'd argue that bitcoin is also an energy hog in theory: Why do people believe there is money to be made in mining? Because bitcoin's core mechanisms ensure that money is to be made, as long as bitcoin is worth enough (which it currently is).
So the bitcoin protocol does three things:
1) Encourage people to mine by offering block rewards and tx fees. 2) Ensure that new miners don't actually add any useful capacity to the network. 3) Require that old miners do at least as much work as new miners if they want to keep their income.
If you take those together, the bitcoin protocol encourages a race to spend more energy to archieve the same thing.
> They will be wrong sometimes, and right sometimes, and equilibrium will be reached intermittently.
There is an equilibrium of energy costs in bitcoin, obviously. However no one knows how the bitcoin/USD price will develop, so I don't see how that implies any cap for energy costa in USD. And even that is just an aporoximation for the actual energy costs in raw materials. One that can be very skewed as the posts further down in the thread explain.
Let people be wasters, nothing wrong with that. If Bitcoin experiment succeeds it's going to be better than the Internet. I hear no one complaining that Internet servers waste magnitudes more energy than Bitcoin, serving porn, FB, flash games and all other wasters of life.
HN is as much your comment as any other. More, in fact, because yours has been upvoted to the top.
Any central bank can use their monopoly over the creation of currency to increase reserves whilst incurring a cost less than those reserves. For example, suppose the US economy grows by 50% over 30 years, but that the country's payment infrastructure doesn't change otherwise, and so the typical American wishes to have a higher real value of cash in their wallet, to buy more luxurious office lunches, or spend the cash in other circumstances.
In a situation where everyone uses a traditional fiat currency, minted by a central bank, the central banks monopoly over the production of these additional currency reserves will allow them to spend significantly less than the value of the currency in making it. Today, a $100 bill costs 12.5 cents to make, plus some additional transport and monitoring costs [1].
However, with the decentralised nature of a crypo-currency, a competitive market dictates that in order for it to be created, miners must spend the entire value of the cryptocurrency in energy costs, otherwise they would be forgoing risk-free profits. Suppose the economy uses cryptocurrency. If the market allows for miners to spend only $90 of energy, hardware and labour to mine $100 of currency, then a less efficient entrant can come in with costs between $90-$100 and make a positive expected profit. This means that even if there are not technical difficulties preventing a crypocurrency from replacing a traditional fiat currency, it's adoption reduces the social surplus generated from the minting of new currency.
[1] https://www.marketwatch.com/story/new-100-bill-costs-60-more...
This author is getting the causation backwards:
> Is there a correlation between power consumption and the cost of bitcoin.
Yes, of course there is a correlation, but he's getting the causation entirely backwards.
The profitability of mining is directly proportional to the price of bitcoin, because you're paid with 12.5 Bitcoins per block. But your costs are fixed in USD -- you pay for electricity costs, per kilowatt. Therefore as the Bitcoin price goes up, it's profitable to run that many more miners, to produce that many more hashes, until you fill up the profit with electricity costs again.
So the price of Bitcoin determines the hashrate, not the other way around, like he's presuming here:
> If this correlation is strong enough, it means the Bitcoin will always be bounded by the cost and availability of electricity.
This guy also assumes that miners are paying the average electricy price in the US and China. Miners pay nothing like that. California electricity rates are around 15¢/kW. Central Washington is 2-4¢. That's 6-7x cheaper. The same is true for China.