The unreasonable ecological cost of #cryptoart
January 21, 2021 3:02 AM   Subscribe

Cryptoart - the category of art related to blockchain technology - has exploded over the course of the last year. This is in part due to platforms such as SuperRare offering 'tokenised' art pieces. These give people the ability to stake verifiable ownership of digital art works, as well as offering artists the ability to produce 'limited edition' digital art. But as Memo Atken found, the ecological damage of this rising art marketplace is staggering.
posted by secretdark (69 comments total) 28 users marked this as a favorite
 


There is someone I have known for many years who has become thoroughly convinced cryptocurrency will take over the world (and also believes in QAnon, etc.). I imagine it is only a matter of time before she discovers cryptoart and starts touting it on Facebook. The more I learn about this stuff the less sense it makes.
posted by TedW at 3:52 AM on January 21 [1 favorite]


Well that's unnecessary. Not literally every implementation of a blockchain involves wasting energy making computers do busy work.
posted by wierdo at 3:53 AM on January 21 [2 favorites]


You could make a similar calculation about almost any aspect of the global economy. Seems to me the solution is not to get rid of crypto but to decarbonize the production of electricity.
posted by Pararrayos at 4:40 AM on January 21 [7 favorites]


Seems to me the solution is not to get rid of crypto but to decarbonize the production of electricity.

whynotboth.jpg
posted by Dysk at 5:00 AM on January 21 [23 favorites]


Seems to me the solution is not to get rid of crypto but to decarbonize the production of electricity.

I concur.

Black markets will always exist. Outlawing maths can really only be performative. If people are using maths to enable black markets, minimising the externalities has to be considered, and is likely the least-shit option.

I wonder if there's a way to attack the confidence in cryptocurrency - lack of confidence being the thing that has generally destroyed traditional currencies, but I can't think of it, and I expect many people smarter than I have been trying this for years.

(plus, y'know, we need to decarbonise the production of electricity anyway, or give up electricity)
posted by pompomtom at 5:12 AM on January 21 [1 favorite]


The article demonstrates the orders of magnitude higher carbon outputs of blockchain activity alongside comparable economic transactions- e.g. the 685kWh footprint (equivalent to 326kg of CO2) for one single blockchain transaction vs. less than 200kWh footprint for one hundred thousand VISA transactions. I don't think saying "oh well, everything we do is bad in terms of carbon output, therefore this phenomenon is not worth criticising" really gets at how terrifyingly energy consumptive this stupid gameified thing is.

Or as the article puts it another way: "If BTC tried to match just 1% of the volume that the global banking industry currently services (i.e. 1 billion transactions / day), it would already consume 30,000 % more than banks do."

The best (?) part for me is how unbelievably bad most of this art is. Like take a look at @beeple, who according to Wikipedia has broken the record for crypto art sales. Or have a flick through the article writer's site http://cryptoart.wtf/which shows the energy consumption for any given piece of crypto art. Or have a browse through Superrare. Even if it was not literally costing the planet, is this stuff really worth it? Of course not, the art itself is not the point, but then that begs the question- why are the images and gifs then being made? What are the reasons for these images' existence? What other symbol could replace them in this cryptoeconomy? (A: pretty much anything, see for example Rare Pepe Wallet

Some artists working in this field, like Robbie Barratt (@videodrome) are doing interesting stuff with AI generated landscapes and figures, and I think it's interesting to think about authorship and creativity in light of these ways of making art. But I think any artist who is actually interested in that sort of thing in a meaningful way of creating a new artistic paradigm which is not stuck in the human-centric mode of being "The Creator"... would steer way clear of this whole monetised/tokenised capitalistic eco-fuck thing.
posted by Balthamos at 5:13 AM on January 21 [20 favorites]


the 685kWh footprint (equivalent to 326kg of CO2) for one single blockchain transaction vs. less than 200kWh footprint for one hundred thousand VISA transactions.

Indeed, but qualitatively they're different transactions because of the whole black market aspect, no? I'm talking more about the idling-car-for-heroin thing than some smart-arse's art project.
posted by pompomtom at 5:17 AM on January 21


I'm still waiting for my etoy art group shares to vest.
posted by loquacious at 5:19 AM on January 21 [2 favorites]


Seems to me the solution is not to get rid of crypto but to decarbonize the production of electricity.

Wasting energy always causes trouble, decarbonized or no.

Seems to me that what we need is more journalists understanding, and then pointing out, that proof-of-work crypto protocols specifically are the ones that waste outrageous amounts of energy when their token price gets high, and that there are newer blockchain approaches (e.g. Ripple, Stellar) based on consensus protocols that are not only faster than proof-of-work but also use energy at similar rates to traditional, non-blockchain transaction processing.

The clear standout in ridiculous energy over-consumption is BitCoin, and this is because (a) it's a proof-of-work protocol and (b) it's been around longer than anything else, so it's had more time for its token price to inflate. And since the amount of energy that BitCoin wastes is directly proportional to the price of its tokens, the way to stop BitCoin burning energy like there's no tomorrow is to induce a BitCoin price crash. And since the price of a BitCoin is defined more purely than that of any other commodity by what people are willing to pay for it, widespread public concern that BitCoin in particular is now a significant contributor to global warming might even be enough to bring on such a crash.

If the whole world starts selling BTC and buying XLM instead, not only will I personally make out like a bandit but we could all avoid maybe a quarter degree of global warming. It's win-win!
posted by flabdablet at 5:27 AM on January 21 [11 favorites]


Black markets will always exist. Outlawing maths can really only be performative.

This is obviously nonsense, no different than saying fire will always exist so outlawing tire burning will only ever be performative.
posted by mhoye at 5:34 AM on January 21 [17 favorites]


Thanks flabdablet - I had thought that all cryptocurrencies were just sort of bitcoin with a different seed or whatever.

I'm sticking with "black markets will always exist" but perhaps if regulators could focus on "Silk road version n"'s that use bitcoin(/other energy-intensive coins) and turn a blind eye to less geosuicidal currencies then the desirable crash might be induced.
posted by pompomtom at 5:38 AM on January 21 [1 favorite]


I wonder if there's a way to attack the confidence in cryptocurrency - lack of confidence being the thing that has generally destroyed traditional currencies, but I can't think of it, and I expect many people smarter than I have been trying this for years.

If the US Government wanted to do so, they could break Bitcoin before noon. They'd just have to put those nuclear bomb simulations on hold for a few hours. Split the chain a few times and there would be utter chaos in the BTC community.

The CCP would find it even easier since damn near half the mining capacity is already in China.
posted by wierdo at 5:43 AM on January 21 [2 favorites]


This is obviously nonsense, no different than saying fire will always exist so outlawing tire burning will only ever be performative.

You've misunderstood my points, for which I apologise.

These are separate statements.

1) Black markets will always exist.

I doubt this is the controversial bit.

2) Outlawing maths can only be performative.

Didn't some US state once legislate pi to be 3? Circles didn't change.
The PM of Australia once pronounced that the laws of Australia trumped the laws of maths, but that doesn't stop public-key encryption working.

It's the combination of these things that seems, to me, to be the issue.
posted by pompomtom at 5:45 AM on January 21 [1 favorite]


Oh, apparently the Indiana Pi Bill was not enacted.

Turnbull is still a moron.
posted by pompomtom at 5:51 AM on January 21 [1 favorite]


If the US Government wanted to do so, they could break Bitcoin before noon.

I honestly don't think that this is plausible any more. Maybe it could have been five years ago, but not now.

The overwhelming majority of BTC mining is currently done using application-specific chips heavily optimized to perform it at the lowest achievable energy cost, and yet BTC mining is currently burning electricity at the scale of small to medium sized countries. This should give you some kind of clue about just how much computation power is currently being devoted to the enterprise. I would not expect the US military/espionage establishment to have instant access to anything like that amount of compute power.

Also the fact that BTC mining ASICs are special purpose and are energy-efficiency optimized should tell you that any attempt to dominate BTC mining using whatever general-purpose supercomputing tech the US does currently possess would, if successful, need to involve burning electricity at the scale of large countries.

Not going to happen.
posted by flabdablet at 6:00 AM on January 21 [5 favorites]


Turnbull is still a moron

I don't think so. But he does play one on TV.
posted by flabdablet at 6:02 AM on January 21 [1 favorite]


I honestly don't think that this is plausible any more.

Perhaps a more organised and powerful country could do it. I bet you they'd save money on shipping.
posted by pompomtom at 6:45 AM on January 21 [1 favorite]


flabdablet, it's not going to happen, but not for the reason you think. It's not going to happen because there are more than a few wealthy individuals who have plowed substantial fractions of their capital into holding Bitcoin.

Remember, it's a 50% attack, not a 90% attack, and the denominator is variable. Plus there are weaknesses in SHA256 that could make such attacks somewhat easier given sufficiently substantial general purpose compute resources. Not to mention the usual CPU miners are shit compared to what is possible because it has been a decade at this point since it was at all financially viable relative to GPUs and ASICs so nobody cares enough to make them more efficient. That changes if your goal is to break Bitcoin rather than to make money and pay a power bill.

Nobody with enough ASICs to break Bitcoin the easy/obvious way wants to do it precisely because it would break Bitcoin, thus tanking the value of their investment. What good is it to take all the Bitcoin if you make it worthless in the process? It's like that early scene in RoboCop where Clarence is pissed because they "burned the fucking money".

It's true that I oversimplified a bit, but not by much and probably not even in the amount of time it would take if fires were lit under the right people's asses at NSA and DOE and they didn't bother with all the bureaucratic bullshit.

Still, the CCP would have an easier time of it since it's probable that there are enough ASIC miners in China that aren't running right now to bring them well over the 50% threshold if they were suddenly reactivated.
posted by wierdo at 6:52 AM on January 21 [1 favorite]


Honest question: who's paying for all this energy?
posted by signal at 6:57 AM on January 21 [6 favorites]


The US does not need to attack Bitcoin on a technical level to crash its value: it could simply outlaw Bitcoin. If the US declared Bitcoin illegal (on whatever grounds), then although that would't stop people from directly using Bitcoin, it would make it vastly more difficult to convert substantial amounts of Bitcoin into or out of US dollars. Since the major attraction of Bitcoin is as a speculation vehicle that eventually turns back into US dollars, I suspect this would not be great for Bitcoin's price.
posted by Pyry at 6:57 AM on January 21 [3 favorites]


The best (?) part for me is how unbelievably bad most of this art is.

I return to a comment I made nearly three years ago which feels just as relevant now as it did then -

Blockchains: is there any problem they can't not solve?
posted by solotoro at 7:05 AM on January 21 [5 favorites]


I had thought that all cryptocurrencies were just sort of bitcoin with a different seed or whatever

The fundamental ideas underlying all blockchains are the same:

1. There is a public ledger with massive redundancy: in theory every participant in a blockchain transaction has a complete copy of the entire ledger, making the ledger virtually impossible to destroy. In practice this kind of redundancy isn't really practical but there are less storage- and network-expensive ways to get plenty close enough.

2. The ledger bundles all its transactions into a series of blocks, where each block contains a code resulting from applying a one-way hash function to the entire content of the previous block. This means that nobody can tamper with the content of any block in the chain (except the very newest one) without that tampering being detected. This effectively makes the blockchain a write-once ledger.

3. There is some method for making sure that nothing nefarious is ever done to the very newest block in the chain, and for making sure that all participants agree on exactly what that block ought to contain.

Part 3 is where the real differences in the various blockchain protocols exist.

Proof-of-work protocols (as used by Bitcoin, Ethereum and many other blockchain implementations) make constructing ("mining") the newest block a thing that any participant in the network can choose to compete for. The successful miner is the first to solve a puzzle involving intense computational work. The amount of work required is constantly adjusted by the network so as to keep the rate at which mining happens roughly constant. Along with the competition and the fact that rewards are paid in the blockchain's native tokens, this adjustment keeps the cost of the energy it's worth burning to mine a block track quite close to the market value of the miner's reward, which is a fixed amount of native tokens. So as the token value is driven upward by speculative forces, the reward for successful mining goes up as well, thereby increasing the amount of electricity worth devoting to the job. And since the market value of a BitCoin is now literally millions of times more than when this stuff was first implemented, what was always an inherently wasteful process has now turned into a global menace.

Proof-of-stake protocols restrict newest-block construction to a subset of participants who can show that they hold more than some minimum amount of the blockchain's native currency and therefore have a stake in preserving the value of that currency. Of these, one gets picked at random to do the deed, then penalized if they don't do it or do it wrong. Ethereum has been on the point of cutting over from proof-of-work to proof-of-stake for some while now, but apparently getting the details right always seems to be harder than anybody thought. I expect commercial fusion power and strong AI to happen first.

Federated Byzantine Agreement protocols let basically anybody do it but impose a carefully designed cross-checking regime on the result to make sure anybody who does it wrong just gets ignored. To the best of my knowledge, Ripple and Stellar are the only cryptocurrencies currently using anything like this. Which is, of course, the reason I bought a bunch of Stellar XLM. I want nothing to do with any PoW blockchain and I can't see any reason why PoS would ever work better than FBA.

there are more than a few wealthy individuals who have plowed substantial fractions of their capital into holding Bitcoin.

Sure, and those people will benefit the most when the Great Selloff kicks off because they'll be the ones kicking it off.

Honest question: who's paying for all this energy?

BitCoin miners. And it's worth their while to do this because the network pays them in BitCoin for doing it, and BitCoin is currently really really pricey.

Of course the ultimate payers for all this energy are the folks who buy BitCoin when it's expensive and sell it when it's cheap; that would be the mums and dads who have recently been reading in all the financial papers about how no it's totally not a bubble or a Ponzi scheme but a serious Financial Instrument because Big Finance is buying it now. Also, Blockchain!

Pro-tip: when Big Finance starts buying a lot of any one thing, that's exactly the wrong time for Little People to start thinking about holding it. The only reason they buy it is because they want to sell it to you for more than they paid for it. Which is, of course, exactly why I want to sell you all this XLM I bought at ten cents apiece.
posted by flabdablet at 7:06 AM on January 21 [22 favorites]


I keep coming back to this. There is this "art"/NFT element that is purely and totally- bonkersly-empty speculation but ties into some really potent stuff about creation and collection, that seems different from just talking about the general phenomenon of cryptocurrencies.

I'm coming from the perspective of an artist who frankly doesn't really get crypto but sure gets that it is very hard to make money from art, especially digital/web art. You can see artists like @beeple making literally millions editioning off their work, and it seems unbelievable. How many web artists even make any money at all? (Or for those who do get famous enough to make money, how many see their best pieces get ripped off and sold as a product by some big company?) But as the article shows, the ecological cost of this golden ticket of money and a meaningful digital "copyright" is absolutely monumentally appalling.

"In 2 months, just for keeping track of bids and transfers of ownership, Beeple’s NFTs have managed to consume the same amount of energy that an average European would consume as electricity, in 90 years. This energy could bring 2500 Olympic sized swimming pools to a boil. The carbon emissions are equal to 230 transatlantic flights."

And as the article goes on to point out, the whole thing is not even actually providing anything that it claims to, it's basically a scam, all you're getting as the purchaser is a certificate which could technically be printed over and over again or attached to different NFTs.

The article writer calls for creators and collectors to boycott unethical platforms. I'm really glad Memo Akten has done the work to collate and publish this data, hopefully pushing a ball down a hill which will lead to artists and collectors disinvesting from these terrible platforms.
posted by Balthamos at 7:17 AM on January 21 [3 favorites]


Thank-you for this interesting topic secretdark and your contributions flabdablet.
posted by Braeburn at 7:18 AM on January 21 [3 favorites]


all you're getting as the purchaser is a certificate

Yeah, you aren't buying an artifact (either digital or physical), and you aren't buying the copyright to a work or a license to it (presumably 'owning' a cryptoart piece doesn't give you the right to sell prints of it, for example). What you're buying is getting to have your name next to the work-- it's Patreon with more steps and orders of magnitude more carbon emissions.
posted by Pyry at 7:26 AM on January 21 [5 favorites]


I followed a handful of machine learning generative art people on twitter around 2016. I've been surprised to see basically all of them migrate to this tokenized art scene and get really into talking about cryptocurrency in the last ~year. I'm totally heartened to see Memo take this on and push back against his peers. Many of them, like Mario Klingemann whom I think has done cool work, really push back whenever he tweets about this and act like the data isn't clear on the energy footprint of blockchain. Hopefully he changes some minds.
posted by little onion at 7:30 AM on January 21 [2 favorites]


What you're buying is getting to have your name next to the work

Seems to me that what you're buying is a thing that nobody has much of a clue about but they're pretty sure they'll be able to onsell for more to the next rube. In this respect it works pretty much identically to the upper reaches of the mainstream art market.
posted by flabdablet at 7:39 AM on January 21 [5 favorites]


Robbie Barrat is an interesting case because he basically got cheated out of a $400,000 "first ever AI" art auction when he was a teenager. Some dudes took a machine learning model he had created to make interesting generated nude paintings (and which was open source), printed one out and sold it at Christie's. He then got rejected from art school. I've assumed he migrated to cryptoart in part because he felt screwed by the mainstream art scene.
posted by little onion at 7:40 AM on January 21 [5 favorites]


wait Beeple is a crypto artist? My reality is being shattered.
posted by Young Kullervo at 8:32 AM on January 21 [1 favorite]


A substantial, progressive tax on energy consumption (with a strong carbon tax component) would be a straightforward way to address the absurd wastefulness of cryptocurrencies.

Or, alternatively, use the ledger against it and impose a cryptocurrency transactions tax, with the rate set to offset the carbon impact of the transaction, plus a pro rata share of all of the transactions for which the tax could not be collected (e.g. pseudonymous transactions between private individuals). Thus, as the blockchain gets longer and mining gets more energy intensive, the transaction tax rate increases.

As a practical matter, the tax would probably only be enforceable against transactions with companies who can be audited, but the fact that cryptocurrency can be used with such companies and not just sketchy greymarket websites is a major part of its value. The more freeloaders there are, the more expensive "real" uses of cryptocurrency become (because of the pro rata share of uncollected taxes), which would devalue the cryptocurrency. The tax would either be something like self-balancing or would precipitate a crash in cryptocurrency value, either of which is fine with me.
posted by jedicus at 9:11 AM on January 21 [4 favorites]


When Bitcoin first became a thing, I was working for a Fintech startup focused on alternative investments. I was curious, but not enough to actually bother mining or buying any. Then, I saw that hedge funds were starting to invest in it, and I immediately noped out of any possible interest in Bitcoin or any other cryptocurrency.

Everything I've learned about cryptocurrency and blockchain since that initial introduction has not inspired any confidence in it being worth investing in or spending any time on besides bashing it whenever it comes up. It's an over-engineered, overhyped piece of shit that techbros insist will be the solution to every problem, 99% of which could be solved by a fucking wiki.

I do own some cryptocurrency that I was given by Keybase, but I've done exactly nothing but spend maybe an hour trying to figure out how to turn it into real money. After that, I decided it wasn't worth my time to do anything with it.
posted by SansPoint at 9:44 AM on January 21 [2 favorites]


> The US does not need to attack Bitcoin on a technical level to crash its value: it could simply outlaw Bitcoin. If the US declared Bitcoin illegal (on whatever grounds), then although that would't stop people from directly using Bitcoin, it would make it vastly more difficult to convert substantial amounts of Bitcoin into or out of US dollars. Since the major attraction of Bitcoin is as a speculation vehicle that eventually turns back into US dollars, I suspect this would not be great for Bitcoin's price.

It would not take much effort for somebody with thousands of dollars worth of Bitcoin to convert to some other currency and then convert that to USD. For those with millions of dollars worth of Bitcoin there are banks that can do it for them. Granted there would be fees along the way, but if the ROI is higher enough that doesn't matter, since it would still be cheaper than conventional money laundering.
posted by at by at 9:46 AM on January 21


"crypto" is short for "cryptography" and I will take on anyone who says otherwise: all at once or one at a time, makes no difference.
posted by rum-soaked space hobo at 9:48 AM on January 21 [7 favorites]


I must be missing something. There's nothing to guarantee that duplicate artwork isn't sold with different names or serial numbers except the artist's word, right? One could invent six artists and create six copies of each original artwork with unique blockchain IDs. So. . . how is this different from handing cash to an artist and receiving a note that says, "this is only going to be sold once," perhaps signed with a personal gpg key? All the legal remedies are the same and there aren't really any non-legal benefits.

If it's to protect the sellers, is keeping track of sales in fine art auctions really a problem that needs to be solved? All the well-known art scams I can think of would be no harder to pull off in this case. Perhaps you'd have better luck trying to prove that someone knew about a thing later, if you assume an incredibly tech-savvy court. But. . . that's about the only advantage I can think of. Why bother?
posted by eotvos at 9:53 AM on January 21 [8 favorites]


A substantial, progressive tax on energy consumption (with a strong carbon tax component) would be a straightforward way to address the absurd wastefulness of cryptocurrencies.

There is enough money floating in BitCoin right now that in the presence of a progressive electricity tax it would pay to set up private utilities to generate energy to power the ASICs. Plus it would be possible to avoid the progression by setting up different utility accounts for each ASIC/computing unit if necessary to keep usage per account under the first step of the price increase.

Also it would be pretty devastating for the aluminum industry.

Also also it would have the effect of encouraging people to use fossil fuels for heating tasks instead of clean electrical energy.
posted by Mitheral at 10:12 AM on January 21


Honest question: who's paying for all this energy?

Everybody. We are all paying for all this energy, in the only currency that matters in the long term. Price tag: one inhabitable planet.
posted by notoriety public at 10:25 AM on January 21 [16 favorites]


since it would still be cheaper than conventional money laundering.

Well, no, you would still need to launder that money. The point of money laundering is to conceal the source of the money. In a scenario where Bitcoin is illegal in the US, you'd need to conceal that your money came from Bitcoin and have an answer when the IRS or whoever asks you where all these yen or whatever you've exchanged for USD came from.

how is this different from handing cash to an artist and receiving a note that says, "this is only going to be sold once," perhaps signed with a personal gpg key? All the legal remedies are the same and there aren't really any non-legal benefits.

"Doing something that can already be done just more complicated" is kind of the crypto world's raison d'être.
posted by star gentle uterus at 10:36 AM on January 21 [2 favorites]


how is this different from handing cash to an artist and receiving a note that says, "this is only going to be sold once," perhaps signed with a personal gpg key?

You can pay an artist three continents away in exactly the same way as you'd pay one three doors down the street.

The most compelling convenience that cryptocurrencies have to offer is that they're worldwide by nature.
posted by flabdablet at 11:01 AM on January 21


Honest question: who's paying for all this energy?

In monetary terms, taxpayers subsidize electricity production, delivery, and purchase of overage. All of us pay the ecological costs, ultimately. Moral hazard pushes the savings onto the cryptoscrippers.
posted by They sucked his brains out! at 11:42 AM on January 21 [2 favorites]


Cryptoart depicting the ecological cost of that piece of cryptoart.
posted by Going To Maine at 12:09 PM on January 21


SansPoint, uh, I’ll take the XLM you got from KeyBase if you’re not going to do anything with it... my user name is the same there as it is here. They might take it back from if you if it continues to go unused per terms. I don’t trust them too much since getting bought by Zoom either. But honestly the easiest way to cash it out for you is probably going to be Coinbase.
posted by floam at 12:30 PM on January 21


Basically we need to cost out resource usage on the basis of how much funding it will require to remediate the damage its usage causes to the commons, and then ensure that charge is directly funnelled into this remediation, and we need to re-evaluate the costing regularly.

In practical terms, this means taxes on energy production which get passed on to the consumers. The taxes need to go into actionable programs with clear, meaningful metrics that enable us to evaluate our impact on the viability of the ecosphere. (Species diversity. Arable land. Global air and sea temperature averages (as a proxy for many other concerns ))

The only way to practically make this work is to both apply this costing at multiple places in the economy (energy production is not the only one - consider for example the global impact of construction), and also ensure individuals have enough resources (concretely in our world, money) to satisfy their needs in this scenario.

Tl;dr:

(A) leverage taxes on major sources of environmental damage as low down in the supply chain as is necessary to implement an effective feedback loop and controls.
(B) reduce the vulnerability of individuals’ quality of life to the effect of the above mechanisms by ensuring everyone has the resources they need no matter how those mechanisms affect the cost of resources from the POV of individuals. This is fundamentally an argument for UBC, socialised health care, public housing, free education, and a plethora of other social programs that reduce the vulnerability of individuals in less fortunate economic circumstances to changes in the costs of basic life necessities.

Sorry, I guess tl;dr isn’t my strongest suit.
posted by allium cepa at 2:43 PM on January 21


I worked really hard to pass a carbon tax in my state. Knocked hundreds of doors.

It failed.

I joked with a friend “America would rather build seawalls than fight climate change.” A few weeks later, I read an article where a Goldman Sachs analyst described seawall construction as “a great investment opportunity”.

Blockchains/crypto are a distraction from the real enemy, which is capitalism.
posted by FallibleHuman at 5:31 PM on January 21 [6 favorites]


> there are newer blockchain approaches (e.g. Ripple, Stellar) based on consensus protocols that are not only faster than proof-of-work but also use energy at similar rates to traditional, non-blockchain transaction processing

Is this really true? Not having studied it I'd expect it would be far better, but I'd think it would be something like proof-of-work is polynomial time with proportional energy consumption, consensus would be linear time, and traditional transaction processing would be scalar time. (Wikipedia: Time complexity)
posted by XMLicious at 7:22 PM on January 21 [1 favorite]


They might take it back from if you if it continues to go unused per terms.

Given that Keybase doesn't have access to plaintext copies of Stellar private keys, I'd be very interested to learn of the mechanism by which you suspect Keybase could remove XLM from its users' Stellar accounts without involving its users.

Seems to me that the closest Keybase could get to nobbling a Stellar account would be to forget the Keybase server's encrypted copy of a user's private Stellar key; perhaps deliberately per some TOU I'm currently unaware of or other corporate malfeasance, perhaps inadvertently by going belly-up.

Personally I guard against this possibility by keeping my own independently secured copies of my Stellar private keys. The only ways I can think of for the XLM I hold to be forcibly removed from my control is if the entire Stellar network falls apart, at which point my private keys become moot.
posted by flabdablet at 8:56 PM on January 21


use energy at similar rates to traditional, non-blockchain transaction processing ... time complexity

I agree that the amount of energy that a FBA protocol requires would be expected to scale with the number of active nodes, though I suspect that linear scaling might be too high an estimate.

The point for me, though, is that it doesn't scale with the market value of the tokens involved, and is therefore not, like PoW, subject to being blown up to catastrophic proportions by market forces external to the protocol design. So as is the case for non-blockchain transaction processing, I can't see the energy consumption of FBA ever becoming high enough to be worth more concern than the energy consumption of, say, Netflix or online advertising.
posted by flabdablet at 9:05 PM on January 21 [1 favorite]


So it's worse than the thing its replacing, but not as bad as some completely irrelevant third activities?
posted by Dysk at 11:22 PM on January 21


allium cepa, can you share that article? I can’t find it using the terms suggested. Like it seems like no such article exists mentioning “a great investment opportunity”, goldman sachs, and seawalls anywhere ever man.
posted by floam at 12:27 AM on January 22


floam, the comment you’re referencing was by FallibleHuman. Mine was just ranting on about the lack of systematic approaches at effective scale.
posted by allium cepa at 4:52 AM on January 22 [1 favorite]


So it's worse than the thing its replacing

Maybe, maybe not. Hard to say. You'd have to look very carefully into how much and what kinds of transaction processing something like Stellar actually could replace to make that call. The analysis is made more complicated by the fact that a single Stellar transaction could potentially replace many traditional transactions in cases where money is crossing national borders. And of course every transaction on the Stellar network that would otherwise have occurred on any of the proof-of-work networks is a huge energy win.

but not as bad as some completely irrelevant third activities?

Not bad enough to be worth worrying about even slightly, given that other services that consume many orders of magnitude more energy than Stellar are uncontroversially regarded as acceptable.

Blockchain has a well-deserved reputation for being ridiculously wasteful, but that reputation is based solely on the way proof-of-work protocols currently dominate the space. Stellar is to Bitcoin as solar panels are to coal power.
posted by flabdablet at 5:24 AM on January 22


If I've got my 10,000-foot view of this right, because again I haven't taken a really close look at anything, I believe the major advantages of the consensus-based approaches flabdablet is describing (when compared with traditional transaction processing) are 1) decentralization and 2) effectively, the decentralization means that the transaction records are very thoroughly backed up. (...and a bunch of other benefits fall out of those first two things.)

In terms of strict efficiency, like comparing engine technologies, the difference in computational resource consumption is still significant; but if the criticism is that this is a frivolous use of computing resources, then switching from polynomial-time algorithms down to sub-linear-time really does bring it so far down below most leisure activities, in scale, that to continue focusing on these lower-power blockchain thingies as a material factor in climate change by themselves would appear to be missing the (on-fire, flaming) forest for the trees.
posted by XMLicious at 8:52 AM on January 22 [1 favorite]


Transaction fees on Stellar are also ridiculously low compared to e.g. Mastercard or PayPal or (especially) Swift.
posted by flabdablet at 9:05 AM on January 22


Floam, it appears I mis-remembered which firm the analyst was from - it was JPMorgan rather than Goldman Sachs. Apologies. See subhead "Build the (sea) wall" here.
posted by FallibleHuman at 10:19 AM on January 22 [2 favorites]


Someone on Twitter asked if they should heat their home with Bitcoin mining. A couple answers (1 | 2) were fascinating demonstrations of late-stage, pre-extinction Holocene thinking.
posted by They sucked his brains out! at 9:31 AM on January 23 [1 favorite]


Yeah, that Bitcoin heating price analysis is either crazily naive or deliberately misleading. Always hard to tell which of these is motivating any given Bitcoin booster.

If there were a way to run a heat pump with a reasonable CoP off the waste heat from a Bitcoin mining rig, that would be different. But the simple fact is that the waste heat off electronics is very low quality (under 100°C) and therefore almost always uneconomical to use except directly.

Any way to divert an existing flow of that kind of low grade heat from waste to good use is of course perfectly fine and usually an excellent idea, but deliberately creating new flows of it where more efficient alternatives exist is terrible engineering practice and very unlikely to perform well.
posted by flabdablet at 12:44 AM on February 12 [1 favorite]


I've installed megawatts of resistive heating. If someone could come up with a self contained, plug and play mining rig that actually paid for itself I could probably sell a lot of them.
posted by Mitheral at 11:43 AM on February 12


Hardware deliberately designed to create a perverse incentive to burn roughly four times as much energy on home heating as is actually needed for that job, especially when a substantial fraction of that energy will probably keep being derived from burning fossil fuels for the equipment's likely service life, strikes me as purely evil.

I'm sure you could sell a lot of those. Would you?
posted by flabdablet at 2:07 PM on February 12


To the extent that resistive heating was going to be used regardless, mining Bitcoin is..not terrible, assuming the use of used equipment, which of course means the financial return will be near zero. I'd rather that computing power used in that way be turned to something more productive, though.

A heat pump would be much better, but they're prohibitively expensive for small space heating uses in most cases. If you need an AC for that space, a mini-split with a heat pump option is usually a small enough incremental cost for it to be financially feasible, but not so much in other situations.
posted by wierdo at 3:40 PM on February 12


I'm sure you could sell a lot of those. Would you?

Instead of a baseboard or a unit heater (and assuming the miner is thermostatically controlled)? Sure. Why not? The current solution is burning exactly the same electricity and getting nothing but heat out of it. If you could also do something else (even if that something else was as useless as bit coin) that offsets the operating costs of the heater why wouldn't you?

There are lots of spaces that need 250-1000W of heat that a heat pump would be crazy over kill for. Especially if that space doesn't have an outside wall. Or it is too cold for an air source heat pump to be effective.
posted by Mitheral at 8:03 PM on February 12


Hey I thought of another use case: hot tubs. While heat pump heaters for spas are available they only work down to -10C or so. In my area 99% of 2-6 person spas are electrically resistive heated and the other 1% are wood or gas fired. I've installed or serviced 100s of hot tubs and haven't seen a single small hot tub using a heat pump. A bit coin miner spa pack would be the s**t. The GPU would even be water cooled.
posted by Mitheral at 8:12 PM on February 12


Sure. Why not? The current solution is burning exactly the same electricity and getting nothing but heat out of it.

Because the heat is very low-grade. In theory yes, you're getting 'free' heat, but you might find yourself having to actually use significantly more electricity trying to keep warm with a lot of very small low-temperature radiator units than you would with a boiler and a pump throwing hotter water at larger rads.

And then there's the fact that it can be more efficient to burn your heating fuel directly, compared to using electric heat derived from the same fuels. It's more efficient to set fire to some natural gas and enjoy the warmth than it is to have a power company burn that natural gas, use the heat to drive a turbine, and then do your domestic heating with the electricity. So to start, you're effectively comparing bitcoin heating to the worst possible alternative and even then, it's not clear that the bitcoin solution is better.
posted by Dysk at 12:01 AM on February 13


A bit coin miner spa pack would be the s**t. The GPU would even be water cooled.

To generate enough heat to keep a decent-sized hot tub cool, you would need so so many GPUs. You might end up having to build a second building, bigger than the hot tub, to house them.

Think about how big resistive heaters can get - like, the heating elements themselves. Now consider that most of a GPU is not optimised or effective resistive heater. There's a bit, sure, but there's only so much you can get out of PCB traces without risking damage to it. Same story for the dies. There's much less of them than a standard resistive heating element, and you can't run them as hard. Oh, and there's a whole lot of bulk that isn't adding to your heating potential because these cards are not designed as heaters.
posted by Dysk at 12:05 AM on February 13


You might end up having to build a second building, bigger than the hot tub, to house them.

Not even close.

Make our hypothetical hot tub round, with a diameter of say two metres. That makes the surface area of the water roughly three square metres. Most of the heat loss will be water evaporating from the surface, and at 40°C that will be happening at about 1kW/m2. So we need roughly 3kW of heating to keep the temperature stable; let's be generous and design for 4kW.

A decent gaming desktop PC will burn something like 500W continuously if run flat-out, so at worst we'd need a box about twice as long, twice as wide and twice as high as a desktop tower. With dedicated mining ASICs instead of standard GPUs, and purpose-built motherboards and cooling loops instead of drive bays and slot-mounted graphics cards, we could easily squeeze that by half.

That said, there is no way on God's green Earth that roughly eight gaming PCs worth of electronics and liquid cooling is ever going to cost less to manufacture than a 4kW heat pump. High purity silicon makes a vastly more expensive resistive heating element than nichrome. Even if the rig ended up mining more money than the cost of the energy it consumed, which it woudn't ever do unless you gutted and upgraded it at least every year, total cost of ownership over installed service life would remain comically higher than for even the shittiest heat pump.
posted by flabdablet at 2:45 AM on February 13


It's more efficient to set fire to some natural gas and enjoy the warmth than it is to have a power company burn that natural gas, use the heat to drive a turbine, and then do your domestic heating with the electricity.

Not necessarily true if the domestic heating device is an electricity-powered heat pump.

But if we're going to talk efficiency, the most efficient way to heat a building is to trap solar heat with high-tech windows and store it in interior thermal mass.

Amory Lovins does this and grows bananas in Colorado. But if you need to retrofit heating to a less excellent building than that, heat pumps should be high on your options list.
posted by flabdablet at 2:54 AM on February 13


A small spa has a 3-4KW element. At -20 and with the lid on it'll run for less than a couple hours a day (the elements are hugely oversized for maintenance duty because of the large losses mentioned above when the lid is off and the water is agitated). Call it 8 KWH per day worst case. With steady instead of cyclic heating that is a 333W continuous. About the same draw as a high end gaming PC at full chit. Obviously all that heat isn't going to be available but then again a mining rig is going to be using more GPUs.

Assuming people are already actually making money mining bit coin (an assumption not based on facts in evidence) then it should be possible to make something like this work, at least in theory.

Certainly in the case of plain space heating where the space heating is better than pure gravy on top of the coin mined because you don't have to pay to otherwise condition the air reducing the cost of your mining operation. In a heavy residential/commercial setting it might be worth using bit coin mining to feed a air-to-water heat exchanger to preheat domestic hot water or process water as well as space heating. It won't be as efficient as a heat pump but it'll be cheaper because of the bit coined mined offsetting input cost. Hell it wouldn't even require any out of the box engineering. Plunk down a cryptomining container next to your mechanical room; use water cooled A/C equipment instead of air cooled; and voila: free water preheat.
posted by Mitheral at 9:41 AM on February 13


Assuming people are already actually making money mining bit coin

The people who are currently making money buying bitcoin are doing it at very large scale on the most energy-efficient ASIC-based mining rigs available, sited in places where electricity is very cheap. These outfits dominate the mining scene to such an extent that the wholesale rate they pay for electricity sets the energy cost to mine a Bitcoin. Anywhere electricity sells for substantially more per kWh than it does for the big miners, which includes every home with a retail electricity supply, Bitcoin mining is not likely to break even except perhaps in occasional short streaks.
posted by flabdablet at 10:13 AM on February 13


A small spa has a 3-4KW element. At -20 and with the lid on it'll run for less than a couple hours a day (the elements are hugely oversized for maintenance duty because of the large losses mentioned above when the lid is off and the water is agitated). Call it 8 KWH per day worst case. With steady instead of cyclic heating that is a 333W continuous. About the same draw as a high end gaming PC at full chit.

Except that, as you say, you need that 3kW peak heat output to make the tub work properly. Hot tubs aren't anywhere near as nice to be in if they start cooling down as soon as you take the lid off them.

And again, I come back to the question of equipment cost. Resistance heaters get as widely deployed as they do because the actual heating parts are dirt cheap to manufacture. That's the economic trade-off for their otherwise ridiculous running costs; the same trade-off, as it happens, that most people agree no longer makes sense when applied to light bulbs*.

If you're in the market for a heating device that costs more up front and less to run than a nichrome resistance heater, I would expect a heat pump to outperform a Bitcoin mining rig by a factor of about five on installation and achieve a lower net running cost on retail electricity and both embed and consume far less energy. You could even dress it up with coloured LEDs if you want to make it look all technical and shit.

*By the way, if your house has resistance heaters then your fancy new LED bulbs will be saving you no energy at all. Resistance heat is resistance heat; your house doesn't care where it comes from, it only cares what the total amount released inside is.
posted by flabdablet at 10:28 AM on February 13


Assuming people are already actually making money mining bit coin

Hi, me from 2011 would like to remind you that calculating mining ROI based on current exchange value may be.....shortsighted.

Sigh.
posted by Sockdown at 7:08 PM on February 15


« Older Thanks for clarifying my understanding   |   "A fever-dream of a hotel" Newer »


This thread has been archived and is closed to new comments