This is an interesting article, but IMHO, its discussions on Hacker News lacks depth. Almost all threads are debating about superficial Bitcoin issues that has already been talked endlessly, and none of the thread focused on the main thesis of this article:
> Bitcoin mining is a complex phenomenon that connects hardware and software, the energy and financial markets. Invisible rules govern every aspect of it. The performance of an individual operation is determined by various external factors that are often hard to quantify and almost impossible to forecast.
It stated that Bitcoin mining as a whole, is a complex social-economical-computational-environmental phenomenon with a huge number of moving parts and uncertainties. The most fascinating fact to me, is the direct impact of climate cycle in southern China on the mining industry, almost with a mythical tone found in some Sci-Fi novels. While Bitcoin is a highly artificial and technical construct, but its yield can be severely affected by weather, some Bitcoin miners even routinely migrate their operation according to seasonal changes like the nomadic groups of the past. The flood seasons are visible on the Bitcoin hash rate chart, and even dominates the date of trade shows and R&D schedule of new mining rigs.
> May-October is the flood season in Southwest China. It is also a festival period for mining businesses as the large supply of surplus hydro capacity significantly cuts down miners’ operating expenses. For small-medium scale miners, the flood season can reduce the cost by as much as 40%. For large miners who own proprietary facilities, the flood season electricity cost is practically negligible. Over 80% of the miners in Xinjiang, Inner Mongolia will migrate in flocks to Sichuan, Yunnan, and Guizhou to take advantage of the discount, and they move back or sell their equipment after the dry season arrives in November.
> Gradually, the industry structured itself around these climate patterns. Like ancient rituals, every year before the flood season arrives, major mining conferences get organized in Sichuan’s capital Chengdu. Some facilities are only open to external customers during the flood season. Manufacturers plan their new product release right before it arrives. Miners race to source the latest and greatest machines in bulk.
It's a tremendous story - and all the details make explanatory sense. But I'm a little skeptical, because if the seasonal effects were big you would be able to see them in the overall hashrate graph at the bottom?
It is a very interesting approach to the aspects which interferes directly to the mining environment. And I totally agree the discussions around this subject normally are superficial and around the market emotions.
Considering climate (weather) aspects is one of the strategic decisions about how to conduct the business in this case.
'Bitcoin is the product of hashpower. The industry wouldn’t exist without incentivizing miners to continuously invest in hardware and burning electricity to augment Bitcoin network’s settlement assurance'
I think this is a common misconception. People don't need to continually invest in mining hardware to keep it going. The entire Bitcoin network could technically be run off a single Rasberry Pi. The network hash rate is only so high because the ROI is there, but without it, people would still be mining.
Case in point, there are plenty of long forgotten altcoins out there that people are mining just for fun. At the end of the day your computer is no different from an expensive space heater. Might as well mine some crypto with it.
> At the end of the day your computer is no different from an expensive space heater. Might as well mine some crypto with it.
Or, you know, you could just turn the computer off while you're not actively using it. It saves you money and does a small little good thing for the environment too.
The other common misconception is that the price of bitcoin is determined by the hashrate. It is actually the other way around, a higher price drives up the hashrate, because a higher price means more demand for bitcoin.
Some founders of altcoins apparently don't understand this and try to buy hashpower to artificially increase the hashrate, hoping to increase the price of their coins. If there is no demand for their coins, an ASIC farm is no different from a Rasberry Pi.
Wait... isn't it worse than that? If I buy enough hashpower to have a measurable effect on the price, shouldn't the price be driven down, by the increased supply? Thus an ASIC farm is worse than a raspberry pi, at least as far as the effect on price goes.
(But, isn't there some majority attack? So maybe I want an ASIC farm as a way to protect my investment...)
Supply of what? The supply increase of coins is constant regardless of the hashrate.
This is exactly why there was no effect on bitcoin price when the reward rate halved. The value of the coin drives mining behavior, not the other way around.
Actually, it doesn't. The difficulty adjusts every 2016 blocks, in theory every ~2 weeks, but if a large percentage of the hash power suddenly disappeared it could be much longer.
For a temporary period of time, before it self-corrects back to ~2weeks. It's easy to leave this part out as it is an edge case that doesn't need the extra sentences qualifying the "normal case" that will be observed 99.99% of the time.
An idle computer clocks down and sips power. Going full bore mining coins results in significant power usage increases. We're talking 10 watts vs 100s of watts.
Right but if your goal is to heat your apartment (and your alternative heat source is also electric) running the computer is a pretty efficient heat source.
Well, efficient is hardly the right phrase. When it comes to heating using electricity then 100% efficiency is on the low side. A heat pump could be many times more efficient.
I just also want to add that gas heating is also effectively >100% efficient relative to electricity because there are no losses incurred on the power generation side.
That depends on how you're making that gas. IIRC algae (the primary ingredient in most fossile fuels) are less than 10% efficient at converting sunlight into chemical energy after metabolic expenditures. (There's also the part where the sun is <<0.00001% efficent at turning fusion energy into sunlight-that-hits-earth, but that problem isn't specific to gas heating.)
What people say is that all electric heat is equally efficient in the sense that all the delivered watts ultimately become heat, but at least in the general area where I live, electric heat of any kind is not efficient in the sense that it is significantly more expensive than burning fuel (natural gas, oil, etc.) at home.
The image of people mining in their residences is the furthest from the truth about crypto mining.
It uses primarily renewable resources or is repurposing pollution directly. This is the only economical way to mine. The vast majority of mining power is done this way, upwards of 80%. How much energy is being used is a purposeful misdirection.
That argument, unfortunately, is a slippery slope. One could list a gazillion things that are good for the environment. The environment argument against Bitcoin has been thoroughly debunked.
There are several valid points of criticism against Bitcoin. I wish more effort is put into it to criticize it along those fronts.
You could pay $384k to hire miners to 51% attack BTC right now (that is, if NiceHash had the capacity). If you cut the hash rate that much the cost would be only $400 or so.
Does it sound like a secure trustless currency if anyone can pay $400 to gain 51% majority?
Correction: pay $400/hr to gain a temporary majority in the block mining lottery. You can do a variety of things with that majority power, but “rewriting history” is a bit misleading. Yes, you can probably get the network to discard some previously accepted blocks, but this only allows you to invalidate your own transactions plus the transactions that depend on them.
Furthermore, any blocks that were mined under high hashpower still need the same hashpower to be re-mined. A one-time reduction in network hash power doesn’t make previously mined blocks easier to mine.
That's 450 BTC/day to buy 50% of the hash rate (as if you even could, that in itself is a ridiculous notion - NiceHash says they have only 300 PH).
Or 18.75 BTC/hour. Or $185,000/hour.
Of course, you're not going to buy 50% of the network's currently running hash power. You'd likely need closer to 100% of the network's current hash power to gain 51%.
Wouldn't you be able to buy goods, reverse the transaction, and buy more goods? That seems pretty significant if BTC is supposed to be ubiquitous (large transactions don't raise eyebrows) and cheap to use (meaning cheap to hash and cheap to takeover).
Depends entirely on the policies of the merchant from which you’re buying goods. I have to admit, I had to read a couple dozen research papers about (theoretical) attacks on Bitcoin before I really understood the nuances. I’m expecting main chain transactions to be relatively expensive, but off-chain solutions can still be cheap. Yes, this requires some trust of a third party, but for small transactions that’s probably unavoidable. Not all transactions are created equal, and trustless settlement is expensive.
Let's not get hung up on details, if you have the resources to consider 51%ing Bitcoin for presumably nefarious purposes, you're probably a government/big finance/oligarch actor and you're probably not penny pinching.
Definitely. I don't really think this hypothetical attack would be possible, I only bring it up to demonstrate the effect that lowering the hash rate would have on the cost of disrupting the network.
1: It's actually, under your proposal, "rewriting a portion of [history] during an attack which costs 400 [not 400 thousand] [dollars] per hour".
2: If you are A (and regardless of whether you're also C) and you sent money to B during the 400$/hr (less than a high-profile lawyer, and probably faster to boot) attack, then yes, changing coins sent from A -> B to A -> C is exactly what you can do.
That's 450 BTC/day to buy 50% of the hash rate (as if you even could, that in itself is a ridiculous notion - NiceHash says they have only 300 PH).
Or 18.75 BTC/hour. Or $185,000/hour.
Of course, you're not going to buy 50% of the network's currently running hash power. You'd likely need closer to 100% of the network's current hash power to gain 51%. So, $400,000/hr.
I mean really, does it pass any sort of common sense test that you could 51% the bitcoin network for $400 per hour?
Exactly, that's (very roughly) why it not only is, but must be, that you cannot, in fact, reduce the hash rate significantly without compromising the integrity of the ledger.
If we cut the hash rate tenfold, which is definitely significant, you'd still need to find someone who has hundreds of thousands of dollars of motivation to perform an attack. Having that motivation is hard when it takes a huge amount of double spending all at once, the attack will damage the integrity of the coin, and clients can be modified to be highly resistant to it.
Settlement assurance is roughly a function of (a) the expected cost of mining a block and (b) the total value of bitcoins moved in that block. In general, the larger the ratio of a to b, the less time a merchant must wait to reach a certain level of assurance.
I haven't read the article, but I believe it will say something like, the continual investment is necessary in order for the transactions to be irreversible (settlement assurance). I'm sure there are some forgotten altcoins out there still chugging along, but it probably does not take much money to hire sufficient hash power to double-spend those alts, or even re-write their history.
For as long as bitcoin miners fight, tooth and nail, to hash incrementally more efficiently than their competitors, it will be prohibitively expensive for most actors to reverse transactions.
Not saying this isn't a possibility, but you'd think in the last 5 years of thousands of altcoins this sort of thing must of happened at least once? I guess the interest has to be there to attack it? Though if the interest was there more people would be mining it. Something like that.
Right, I guess I feel less at risk of being the victim of a double spend and more at risk with rewriting history. Are there any reports of that? Like coins I've held on the chain for some time being stolen?
A double spend is generally what people mean by "rewriting history".
A 51% attack only lets people steal money by taking back recent transactions and sending them elsewhere. Transactions that have been received a long time ago are generally safe. (An attacker can reach further back in time if they run the 51% attack longer or with even more hashpower.)
51% lets you rewrite arbitrarily-deep history, limited only by how much time you're willing to spend on the attack. You have more hash strength than the rest of the network combined! You pick some ancient block and start building blocks off of it, and the very definition of a 51% attack is that your new fork will eventually outgrow the "real" fork.
Double spends are possible well before 51%. Even a 10% attacker has the ability to (sometimes) double-spend. A 10% attacker can try to double-spend transactions and has a certain chance of being able to. The chance of success decreases exponentially, the deeper in the chain the target transaction is. However, for a high-enough value transaction, attempting to double spend it can have a higher EV than the mining revenue you lose out on by mining honestly.
The value of your long-held coins will be almost nothing if the network is vulnerable to double spend.
The potential losses are huge (especially for exchanges etc.) from double spend attacks, and the value of a cryptocurrency is very much dependent on users' confidence in its resistance to attacks.
BCH does have a significant amount of hashrate. A 51% attack being possible is one thing, it being profitable is another.
If it's not profitable, it is unlikely to happen. If Hashrate was to drop significantly without commensurate price drops, attacks will happen and then the price will drop to compensate.
If you sell something in return for coins, and the network could agree on a fork of history in which you never received those coins, not all your balance is safe. Either you trust the network operator(s) not to accept a fork, or you trust the implementation to make a winning fork infeasibly expensive (51% of all miners).
Bitcoin Gold was hit by double spend multiple times, and that's just what I remember off the top of my head.
However there are obviously diminishing returns for such an attack. The alt coin value drops drastically when users find out about the attack and flee to more secure currencies.
The point of the numbers is to prove that you heated space. Seriously. It proves that you burned value that you know you are not getting back if you are caught falsifying records for the system. And, everyone knows its extremely hard to not get caught if you try. Therefore you can be trusted to sign records for the system because you are risk:reward tremendously more financially motivated to be a good actor than a bad one --including whatever nefarious scenarios you might imagine.
i understand the idea of proof-of-work. but idk, when i think about all that electricity and computing power wasted on guessing arbitrary numbers... just doesn't feel right
kind of reminds me of how FAANG job interviews require memorizing not-very-useful algo stuff to weed out people who "aren't motivated enough" to study that stuff for the interview¹... "proof-of-motivation" if you will. it serves a purpose, but requires "burning" a lot of ultimately pointless effort
---
1. at least that's what people say about them, never interviewed there myself
"Pointless" only matters if it is "a lot" and "a lot" is relative. It seems like it has not been possible to resolve the question of whether the effort is necessary. If you suggest that existing monetary systems have orders of magnitude lower overhead, particularly fiat money, then people will argue that all sorts of costs are hidden. But I don't believe it makes up the difference, and I attribute peoples' willingness to believe it to the assumption/belief that anything that virally spreads cannot be drastically worse (in some dimension).
That said, IMO Proof of Work is indistinguishable from Proof of Stake with extra steps. And considering the massive environmental externality of Bitcoin mining, I want to see the major nations heavily tax, if not completely outlaw, Proof of Work mining.
> The entire Bitcoin network could technically be run off a single Rasberry Pi. The network hash rate is only so high because the ROI is there, but without it, people would still be mining.
That's a bit of a naive view. It's like saying, we don't need this expensive security vault in our bank, we could just store it all in a shed.
the proof of work is crucial to the proper functioning of bitcoin, and if the only miners were running on a raspberry pie anyone could come along and do a >50% attack stealing what they want.
> At the end of the day your computer is no different from an expensive space heater
This has been false for probably a decade or more. Pretty much any modern processor has dramatically lower power consumption when not under load through a combination of things like clock adjustment, powering off cores, and even dynamically adjusting voltage. This is part of why smartphones have battery life measured in dozens of hours instead of 1-4 hours.
> The entire Bitcoin network could technically be run off a single Rasberry Pi.
But can that be done practically? At some point you lose the ability to tell which of the forks is 'Bitcoin' and which are 'pretender' altcoins.
I can imagine a future where you run a chain for 6 months on a Pi then someone turns up with an ASIC and writes a new 6 month history and calls it Bitcoin instead. The risks of shenanigans would be huge.
In order to maintain network stability at current prices of Bitcoin, a massive amount of hashrate is required.
Those altcoins that have become worthless and are mined "just for fun" are only protected by the fact that they are worthless. Any attack would be a waste of resources. You could still attack them "just for fun" of course...
That's kind of the point, you can back write a block chain with enough hash power. Just go back a few hundred blocks in the chain, and "fork" into an alternate chain. If your chain can catch up and become longer, you win.
This happens on a micro scale every day, when two people "solve" the next block at the same time. Inevitably, one of those forked chains gets longer and the other one is abandoned or "orphaned".
Doesn't the hashrate need to stay high for security reasons? If miner participation is low, can't bad actors dominate the system and produce false transaction?
If the hashrate ever collapses isn't the whole system at risk?
Sorry for not being clear, I meant when all 21 million bitcoins are mined. The transaction fee is pretty small compared to the reward from the new coin, right? Also, since many cryptos have addressed the incentive issue. Maybe let's just focus on bitcoin specifically?
Fantastically written article, very to the point without too much flourish, and very in depth as others have noted.
This bit is funny imo:
But in late 2019, the National Development and Reform Commission (NDRC) removed mining from a list of activities to be eliminated. Note that the NDRC also oversees the energy industry. Massive amounts of hydropower squandered during flood season is a longstanding issue. The officials are beginning to realize that Bitcoin mining is a highly effective way of transforming excess local capacity into a global digital commodity.
One point of view would be that speculation and greed of bitcoin hodl-er fanboys is subsidizing China's renewables industry...
This analysis misses altcoin mining. When mining adjusts on BTC, miners don't just turn off their rigs. They can go mine altcoins. This behavior introduces another dimension to mining economics. Miners are practical. They have to make money with their hardware and electricity. They can't afford to be ideological.
In the Bitcoin mining world (SHA256), miners like to have BCH, BSV and other alternatives. Altcoins allows them to diversify. That's why these chains will never die. They always have mining support.
They'll die if they can't make enough returns to pay for the electricity (and I'd be happy if we helped make that a reality through a tax on carbon and/or POW mining itself).
Mining revenue is the network cost of running the Bitcoin network.
What if you ran a business that spends billions per year on network cost, then somebody told you they could reduce it by 98% and make it a flat cost, forever. That's proof of stake.
Ethereum launches the first phase of proof of stake this year. When Ethereum v1.5 launches in ~18 to 24 months, Ethereum's network cost will undergo such a transformation. The millions per day paid to Ethereum miners will stop being paid, forever. It's a ~98% cost reduction in perpetuity.
Proof of stake is as equally expensive as proof of work. The trick is that both of based on burning money, proof of work burns by using CPU cycles while proof of stake burns through lost interest.
The problem is that the claim that nothing is as cheap as proof of work as made in that article is only applicable if you think cryptocurrencies have value in a way that is comparable to tangible goods in reality, like the electricity they burn to have proof of work.
I don’t. Your “lost interest“ on a currency that has nothing beyond speculative value is literally worthless to me. The increased energy production and subsequent pollution that proof of work costs, however, has a real value and does real damage to the environment.
Proof of work currencies need to be shut down, for environmental impact reasons if nothing else. You keep playing with your speculative assets all you want. I don’t subscribe to the idea but it doesn’t bother me, until you start damaging the real world with it. Then I have to insist you stop.
You can have whatever values you want, but it remains a fact that cryptocurrencies can be sold for real money and they can be lent at interest using "DeFi" so forgone interest is a real thing that can be calculated.
To bring $1 worth of US currency into existence, it only costs a fraction of that amount.
But to bring $1 worth of bitcoin into existence, it costs $1.
So, it seems obvious to me hash based currency is a virus that is many orders of magnitude less efficient, but many people seem impossible to convince of this.
> To bring $1 worth of US currency into existence, it only costs a fraction of that amount.
> But to bring $1 worth of bitcoin into existence, it costs $1.
But this just says the seigniorage value of creating bitcoins is 0. It doesn't say the total value of creating bitcoins is 0 -- they might continue to produce value after being created!
For example, the situation you describe applies in full to actual physical currencies. The Somali shilling trades at the (quite low) cost of printing the paper note. But the existence of the shilling still has value; it enables commerce in Somalia.
Economies operating on a basis of metal currency, or cowry shell currency, also have this feature - pretty much by definition, the trade value of the currency is mostly just the cost of producing it. (For coins, the trade value is usually slightly higher.) But every time a trade occurs, gains from trade are produced, so the currency constantly generates additional value just by being in use.
I wasn't confident enough to declare bitcoin is equivalent to, say, gold. But it sounds like you are, so maybe that is correct - bitcoin is just as bad as gold, or conversely, gold is just as bad as bitcoin. We observe that fiat currency has mostly driven out metal currency, and I'm inclined to think that is because it is more efficient in using resources.
As you say, currency constantly generates value. It is useful. But -100% of the face value is a large burden to start with. How large, relatively speaking? Well, I don't know exactly what to compare it to, but intuitively the value of currency would be related to interest rates, which have been really low for quite a while in the developed world.
Nobody would be mining any crypto if their cost was exactly equal to their revenue. This fundamental misconception might be the reason why you can't convince anyone that crypto currency is a "virus".
In any efficient market, profits approach zero. I don't see the relevance of saying "well, actually, there is some profit". That profit is the mechanism that drives the mining cost to match the value of the currency, isn't it? Please explain more about what you think I am missing, because it sounds like you are just referencing the mechanism that causes the problem.
The consensus part of 'Proof of Stake' systems is crazy complicated, and there really has been no simple way of doing it determined yet. It's still very much an experiment.
Proof of work on the other hand is an order or magnitude simpler, and has worked for the past 10 years and is pretty battle hardened at this point with everyone trying to attack it. I don't know how well the 'consensus nodes' would fair under the same pressure.
> The millions per day paid to Ethereum miners will stop being paid, forever.
Do you mean paid _by_ Ethereum miners?
The network pays miners rewards (which will continue under proof of stake, ie the cost _to the network_). The miners pay the cost of running their mining (eg electricity and hardware costs), which is what will be reduced by proof of stake.
>"If mining operations migrate to more diverse locations in the future and become less concentrated on a single power source, this cycle will cease to play such a significant role."
Why should that happen? After the initial very diverse mining, it only ever got more concentrated to places with cheap energy. The place my change but there will always be an optimum place.
"The more transparent something was, the more mysterious it seemed. The universe itself was transparent; as long as you were sufficiently sharp-eyed, you could see as far as you liked. But the farther you looked, the more mysterious it became." ——Liu Cixin "Three Body Problem"
> Bitcoin mining is a complex phenomenon that connects hardware and software, the energy and financial markets. Invisible rules govern every aspect of it. The performance of an individual operation is determined by various external factors that are often hard to quantify and almost impossible to forecast.
It stated that Bitcoin mining as a whole, is a complex social-economical-computational-environmental phenomenon with a huge number of moving parts and uncertainties. The most fascinating fact to me, is the direct impact of climate cycle in southern China on the mining industry, almost with a mythical tone found in some Sci-Fi novels. While Bitcoin is a highly artificial and technical construct, but its yield can be severely affected by weather, some Bitcoin miners even routinely migrate their operation according to seasonal changes like the nomadic groups of the past. The flood seasons are visible on the Bitcoin hash rate chart, and even dominates the date of trade shows and R&D schedule of new mining rigs.
> May-October is the flood season in Southwest China. It is also a festival period for mining businesses as the large supply of surplus hydro capacity significantly cuts down miners’ operating expenses. For small-medium scale miners, the flood season can reduce the cost by as much as 40%. For large miners who own proprietary facilities, the flood season electricity cost is practically negligible. Over 80% of the miners in Xinjiang, Inner Mongolia will migrate in flocks to Sichuan, Yunnan, and Guizhou to take advantage of the discount, and they move back or sell their equipment after the dry season arrives in November.
> Gradually, the industry structured itself around these climate patterns. Like ancient rituals, every year before the flood season arrives, major mining conferences get organized in Sichuan’s capital Chengdu. Some facilities are only open to external customers during the flood season. Manufacturers plan their new product release right before it arrives. Miners race to source the latest and greatest machines in bulk.