Hacker News Reader

Comment by kryogen1c | Hacker News Reader

> responsible for $0.49 in health and climate damages

red flag this article is steeped in pseudoscience.

you could spend a lifetime trying to model and predict global weather patterns _with no outside intervention_. the idea that you could even make an attempt at attaching a number to this phenomena is laughable. what are the error bars here, 10,000%?

lets pretend, for example, you tried to estimate the impact of tsunamis assuming that we think this weather event is standard deviations more likely due to BTC (this is non-obvious and probably not true, but for the sake of argument). dozens of tsunamis around the globe will kill people and decimate cities, but your estimate will be multiple orders of magnitude off in some places like fukushima - a cursory glance at wikipedia indicates a lower bound of 300 billion dollars.

this is not to say that we shouldnt be concerned about rapid global climate system changes; on the contrary! junk science like this actively harms the cause. we dont have to sensationalize, reality and truth are scary enough.

View on HN

You clearly haven't actually looked at how the damages are modelled have you? You just in your head inveneted a naive way in which to estimate damages and then argued that that doesn't work.

There is a large and robust literature to assess these type of damages (search for Social Cost of Carbon). The methodological challenges are discussed really well in the paper actually. Much of the damages calculated here aren't even climate but direct air pollution, for which we have robust data and damage models. Insurance companies use these type of damage models at large scale for risk assessment all the time.

The article looks fairly solid to me.

The article is undermined by its arbitrary focus on crypto. The premise is that crypto mining drives consumption for power, and power production results in "health and climate damage". Are we going to police consumers of power now? Will we invest our time and energy in determining the "good" and "bad" consumers of power. Or are we going to invest our time and resources in building out a sustainable, clean, and efficient power production infrastructure to replace the ancient one we have?

Ctrl+f 'nuclear': 1/1 found.

Yes, reducing energy waste is an important way to tackle the environmental externalities of energy production. For example, in cold countries one of the major efforts for helping with climate change is to encourage buildings to use better insulation, to waste less heat during winter.

This is right. It is useful to know how much electricity production costs in health and environmental output. It is useful to know how much of bitcoins value is the electricity cost. We can put a value on environmental harm caused by total electricity usage. But, using these and suggesting that we can accurately measure the cost of cryptocurrency now and in the future compared to how the money/value would otherwise be spent is a distortion. The same reasoning can be used to make anything that stores value or does not have an obvious output look disproportionately harmful. It could be useful to compare cryptocurrencies with themselves. I am a little interested to see how they factored cryptocurrency's optimization to use the cheapest electricity possible. It disproportionately favors storing value from electricity that would otherwise be wasted.

If you're not just reacting to the focus on crypto mining, but actually interested in the answer to your question, then the answer is yes.

In fact, energy efficiency[0] is a large area of focus, especially with regard to ongoing climate change[1]. You might have noticed a shift in things like light bulbs building standards and so on. We've made progress[2], and so it is really hard for people celebrating that progress when they see a sudden boom in energy use to create wealth for a very tiny subset of people, a boom in use that threatens to undo all of our progress and then some.

0. https://www.energy.gov/science-innovation/energy-efficiency

1. https://www.weforum.org/agenda/2019/09/why-investing-in-rene...

2. https://www.nrdc.org/experts/sheryl-carter/ramping-energy-ef...

>a boom in use that threatens to undo all of our progress and then some

Then we're back to policing people's energy use. There will be little ROI in going down this path, and a huge opportunity cost.

This is all not to mention the incentives driving crypto. We have many smart minds engaging in this realm - imagine if those same minds were allocated, via a better economic system, to improving our energy production and distribution. Addressing that dilemma is real policy. Unfortunately, in the West we play whack-a-mole and end up with policies to address the effects of prior policies.

The worst kind of criticism about scientific papers is when people say "but the paper that didn't study this topic, didn't study this topic!"

Would you expect a paper on COVID mortality also discuss heart disease and ALS? Of course not.

>Would you expect a paper on COVID mortality also discuss heart disease and ALS? Of course not.

Power consumption and power production are simply directly related. Your clumsy attempt at an analogy here really misses the mark.

The root cause of this issue is our power production isn't clean enough, and therefore consumption drives pollution. What do you address first - the demand or the supply?

It's quite simply amazing to see so little ability to grasp anything beyond immediate consequences of a problem, particularly on a site like this.

Perhaps you are extending the discussion far past the point I was trying to make in a single sentence.

Discussing things with people on hacker news is frustrating because people like you extend the argument to way way way past the point at hand. All of a sudden we are talking about clean energy deployment, not that BitCoin mining is using a lot of power and resources.

The paper has no perspective on electric energy consumption of other activities, there are many more with similar or bigger consumption (ad industry, MIC, etc.) and still advocates regulating specifically cryptocurrencies. The window of opportunity, attention span and resource for effective regulating economic activity is limited. We should focus on the (few!) important polluting classes such as transport and power production, not hundreds of minor consumer classes such as cryptocurrencies or aluminum producers. Focusing on minor issue, cherry-picking contorted arguments and advocating targeted policy is not how respectable scientific paper should look like.

>We should focus on the important polluters, not minor consumers

Do you feel the same way about causes of death?

Imagine saying "who cares Cystic Fibrosis kills x people, heart disease is worse!"

Also, other consumers like cement, fossil fuel production, and plastics HAVE been focused on. Bitcoin is new and only recently popular so we don't know much about it and it's true cost.

>Imagine saying "who cares Cystic Fibrosis kills x people, heart disease is worse!"

You're completely missing the point here: we have finite time and resources to deal with problems. We must dedicate these finite time and resources in the most efficient way to possible to achieve the most desired outcome.

If by "worse" you mean "heart disease causes more deaths and drives higher social costs" then yes there is a strong argument to be made on addressing the root causes driving heart disease with more resources and time then cystic fibrosis. But, again, your analogy is contrived-and-forced-apples-to-oranges with the topic at hand.

This paper is just saying there are reasonably estimable costs. That's all!

It's not even trying to address the bigger argument you're talking about.

>This paper is just saying there are reasonably estimable costs. That's all!

Wrong. You make numerous comments in this thread asserting that the sole focus on this paper is quantifying externalities of crypto mining when that's plainly false. The paper is prescribing policy options for dealing with crypto externalities and birthing new loaded clinical terms to ascribe solely to those externalities: "cryptodamage". Read it.

"As an alternative to intervening in emergent cryptocurrency markets through prices (e.g., taxation) or outright restricting their development, there may be a role for government investment into R&D that focuses on designing mining puzzles that greatly reduce energy consumption (and thus production costs) while still allowing for secure validation of anonymous transactions. Fully vetting and encouraging such currency alternatives might retain the libertarian social benefits of the blockchain [35] while making “the last Bitcoin” irrelevant. Although there will still be private benefits captured from mining under possible low-energy design alternatives, as illustrated here there would be significant public good benefits (reduced cryptodamages) to justify such government investments into R&D. Importantly, such investments in R&D are not mutually exclusive with taxing to internalize cryptodamages, and may also come with the potential to generate spin-off benefits from blockchain technology more generally."

I did read it, before I made "numerous comments."

Papers discuss the implications and applications of their work. That discussion is not the work itself; it is meant to give a starting point for fitting the work with other research or doing new research that builds on the work. Thus, the authors "close by sharing our brief thoughts on policy implications of recognizing cryptodamages." That section mentions a wide range of policy implications—several paragraphs that you didn't bother quoting—without advocating for anything in particular beyond reducing harm.

Frankly, that leaves it sounding like what you're really bothered by is that the paper points out that there is harm, and advocates for reducing it. And if that's the real problem—that you don't like it when someone speaks up and says hey, this has quantifiable downsides—just be honest about it.

> That section mentions a wide range of policy implications—several paragraphs that you didn't bother quoting—without advocating for anything in particular beyond reducing harm.

Yes but that is still propagating the unsubstantiated view that cryptocurrency mining is an important cause of societal harm worthy of mitigation/regulation. There is no comparison to other similar or bigger economic activities, no mention of the fact that the pollution happens at power plants and is under direct control of the power plant owners, not miners. It is inordinately focused on societal harms of cryptocurrency mining, it ignores the benefits beyond mining profit(which is uncertain but non-zero and hard to quantify in dollars). Limitation of the paper narrative to such a narrow topic shows bias.

Not doing what you'd personally prefer isn't a bias. Tackling a problem at several levels is perfectly legitimate; it's just that a paper about clean power generation would be a totally different paper.

It's a single paper. One piece of information that's small enough to be digested and connected with other work. And frankly, if the result had been that each $1 created 1¢ of costs, you'd be celebrating it. Your only real criticism here is that you don't like the result.

If you're so concerned about the root cause of power generation, go focus on solving that problem.

>If you're so concerned about the root cause of power generation, go focus on solving that problem.

So ignore the root cause altogether, lol. And when we've banned crypto mining and accrued even worse social costs as a result, we'll rinse and repeat this process with the next ".*damage" power consumer. Totally sane!

The irony of this whole chain is that the paper itself actually acknowledges this other variable (that y'all are pretending is the only variable) directly:

> There is considerable evidence of concentration of mining operations in particular locations, typically where reliable electricity is cheaply available, though, precise data are lacking (e.g., [15], [6]). In the US, perhaps the most well-known concentration is the Mid-Columbia Basin area in central and eastern Washington State, where cheap electricity is produced by hydropower along the Columbia River [4], [6], however, mining in other US locations also occurs (see [11] and news references in [23], [24], and [25]).

> emissions rates per kWh will differ substantially given underlying differences in how power is produced—i.e., China relies extensively on coal power (>60% of electricity generation) while the US is more balanced (32% natural gas, 30% coal, 20% nuclear, etc.).

Saying they ignore this is just a lie. Criticize the paper, by all means, but this simply isn't a hit piece. It's careful in stating exactly what it can commit to, and it reasonably tries to explain how that fits into the bigger picture.

>banned crypto mining and accrued even worse social costs

Could you describe what social costs we'll have by banning crypto mining?

Banning anything that's already widespread implies making a part of the population into criminals. With that comes the costs of enforcement and of government's authority (which is finite). Consider too that crypto is, often, a response by many to perceived policy for fiat.

I'm sure there is a way to turn off phase out Bitcoin without making thousands of people criminals.

The planet is burning, the cost of enforcement will be worth it.

> Do you feel the same way about causes of death?

Yes? Polluting power plants kill many more people than the whole bitcoin mining operation.

> Also, other consumers like cement, fossil fuel production, and plastics HAVE been focused on. Bitcoin is new and only recently popular so we don't know much about it and it's true cost.

Okay, we agree on something here. I fully support assessing Bitcoin mining and operation costs and comparing those costs to other consumer classes to get perspective. When we do this, it is easy to see bitcoin mining is a non-issue climate-wise (for now). The main issue is pollution done by Big Polluters such as naval transport and coal/gas power plants.

These bogus calls for regulation/shutdowns of minor electricity consumers just spoil the climate discussion, distract us from the real polluters.

Are there parties that benefit disproportionately by establishing a cost for carbon and the definition thereof?

Is this all an altruistic exercise in saving the planet?

If so, it sounds very nice. We are lucky to have the concern of such well funded saviors.

Ah yes, the "this is obviously a conspiracy" defense despite offering no evidence of such a conspiracy at all.

I've posed questions about incentives.

In lieu of engaging in a discussion, you've dismissed the possibility. You've provided no rationale (or evidence for that matter) for this flippant dismissal. Instead, you offer the 'conspiracy theory' pejorative. It is not a starting point for a conversation.

I have not dismissed the possibility there is a conspiracy.

I'm only calling out that asking a leading question like this doesn't add to the conversation at all.

There is an incredible irony to accuse me of providing no rational when this whole thing is about somebody alluding to a conspiracy with no rational at all.

The starting point to a conversation should be "hey, there is a conspiracy here, this is why."

>Are there parties that benefit disproportionately by establishing a cost for carbon and the definition thereof?

Asking about incentives, thinking critically about conflicts of interest doesn't add to the discussion because that might "allude to a conspiracy"?

I don't know if I agree with that. For me it makes sense to first ask "Cui bono". Establish a motive and then search for evidence to prove a hypothesis.

Posing the question is simply the beginning of the process. If you're closed minded about opening the door, my (reasonable or unreasonable, open for debate) expectation would be that you would dismiss anything within as "conspiracy theories". Hence the discussion becomes futile.

I read your reply as skipping over the possibility of a conflict of interest in research rationalizing a carbon credit trading regime.

Saving the planet is hardly altruistic, after all you are living on that planet yourself. So in essence it's at least partially driven from a position of self-preservation, which is by its nature egoistic.

> There is a large and robust literature to assess these type of damages

even granting you that we have a robust model to assess damage (which on its face seems to be a wildly improbably claim) we do not have an accurate model that predicts WHAT damage will be done due to humanity's impact on weather change, so the resultant-cost benefit analysis still propagates that uncertainty forward.

> (search for Social Cost of Carbon)

i havent read it all because i just looked it up[0], but it directly contradicts your claim of our "robust data and damage models". section 5.1 "damage functions" is about how previous damage models are all incorrect.

side note: one of the paper authors is in references 30 times. idk if this is common or not, but it seems suspicious.

> The methodological challenges are discussed really well in the paper actually. Much of the damages calculated here aren't even climate but direct air pollution

wouldnt know, dont have access to anything but the abstract.

look, im not well read into the field of future predictive damage modeling. you may be right that the field has some definition, however i stand by my smell test. even ignoring everything relevent about health, damage, and climate, the start of the sentence "$1 of BTC" is enough to give pause. BTC price is up 30% in the last month and 470% in the last year. not exactly rigorous scientific measure.

[0] https://www.nber.org/system/files/working_papers/w28472/w284...

Bitcoin mining effort is proportional to the cost per bitcoin times the number of bitcoins awarded per block. Since the number per block is constant over each four-year period, it makes sense to talk about the cost per $1 of BTC.

BTC proof of work doesnt determine the BTC/$ conversion. the price is non-scientific, highly speculative, and highly volatile.

The causation works in the other direction, due to the difficulty adjustment.

BTC/USD exchange rates are at best correlated with mining difficulty and 100% not causally related. this is patently false.

That's the reason why the paper is focusing on damages per dollar of mined bitcoin, instead of damages per mined bitcoin.

Please omit personal swipes from your HN comments, no matter how wrong someone else is or you feel they are. Your comment would be just fine without the first paragraph.

https://news.ycombinator.com/newsguidelines.html

Unfortunately I am to late to edit my post. I understand the intent here, but these threads are full of what are, in the end, drastic accusations of scientific misconduct (or raging incompetence) against the authors, made with confidence from a place of proud ignorance. I believe it is important to point out when arguments are not made in good faith. Responding to them as if they were honest attempts at debate is problematic in the sense of a false balance.

That's actually against the site guidelines, which ask you to "Assume good faith."

The reason is that it's hard if not impossible to accurately assess other commenters' intent, and there's a strong default on the internet to assume bad faith. We all need to consciously counteract that if we're to have substantive, thoughtful discussion.

It's true that some commenters behave abusively in ways that make assuming good faith impossible, but (a) I doubt that was the case here, and (b) the guidelines ask you not to reply in such cases, but rather to flag the comment and/or email hn@ycombinator.com.

https://news.ycombinator.com/newsguidelines.html

This is not just some article, it's a peer-reviewed study from the University of Arizona. Surely it has fault, but honestly I don't think a "cursory glance at wikipedia" provides a compelling counterargument

>This is not just some article, it's a peer-reviewed study from the University of Arizona.

When junk science is peer-reviewed, it's even more of a problem.

Sure, but how do you tell that this is "junk science" as you call it?

Perhaps it disagrees with their personal worldview, because it would make their Bitcoin investment a questionable move for the world. There's a lot of justification going on around people's bitcoin investments; you can't say a negative thing about it right now without being dogpiled.

Bad arguments. Estimating social benefit of cryptocurrencies by the dollar cost of mining operations is obviously incorrect.

No perspective on the big picture. Pollution they talk about is happening at power plants, not mining rigs. Net electric energy consumption of cryptocurrencies is less than 1% of world total.

Proposing targeted regulation policies on minor consumer classes, no mention of big problems(transport, coal plants) and big solutions (nuclear energy). Looks very biased.

This looks like a nice lobby/news propaganda scribble from a think tank working against cryptocurrencies. Not as a scientific paper.

> This is not just some article, it's a peer-reviewed study from the University of Arizona

Without going into the paper itself, why do you believe being peer-reviewed means that we cannot - and should not - look at it critically with our own brains and minds, analyze the evidence, and reach our own conclusions?

The comment did not say anything like that. You're making a strawman.

I would argue that most people actually cannot, at least not without spending weeks reading up on the topic first. Needless to say that the same goes for almost every other scientific article out there, except for the ones that might be in one's own field.

University of New Mexico? (clicked 'Show More' under author names - maybe I missed a reference?)

Did you actually read the article? I'll make it easy for you: https://scihubtw.tw/10.1016/j.erss.2019.101281

They make no attempt at trying to predict global weather patterns. Rather the article seems to be mostly "glue" -- connecting energy usage of cryptocurrencies with previously done work trying to measure the externalities of pollution and energy usage. The research that the article builds on gets used all the time in government policy etc. Which doesn't make it correct, but deserves a little more than a knee-jerk reaction and outright dismissal because the method you came up with in thirty seconds doesn't hold up.

This doesn't engage with the report at all. You call it pseudoscience without any supporting evidence. You speculate about error bars without grounding in fact, and the only counterargument you bring is a weak hand waving that it must be a hard problem, with a gratuitous citation from Wikipedia.

And given that you basically claim that actuarial science is impossible, I'd hope you bring a stronger claim than that.

As it is, this is, as the quote goes, "just your opinion, man".

Hyper competitive market forces in the west have created the wrong incentives for everyone, especially institutions. When there is a demand to weaponize or politicize science then we naturally get stuff like this.

You could make up some fear mongering "study" about almost every energy using topic. AI and video rendering for instance which is often for some marginal purpose. Or heating pools, idle electronics, big homes with AC, leisure travel, all the useless crap bought and manufactured in China with dirty coal, etc.

Right, and all those things are harmful to the planet.

Many I just listed use more energy than btc...so yes, they are.

So what are you saying, that because other things are also bad we can't have a discussion how bad one thing is?

Im saying you were being sarcastic thereby making the claim that those things are not harmful, and they are. Im not sure whats confusing here.

I was not being sarcastic. Those things are harmful.

You mentioned five things, all of which are frequently focused on as contributors to global climate change. Well, I'm not sure I've seen "AI and video rendering" mentioned in that context, but I'm also not sure that they use more power than crypto mining. But definitely I've seen articles addressing idle electronics[0], big homes with AC[1], leisure travel[2], and "all the useless crap bought and manufactured in China with dirty coal"[3].

So it shouldn't be surprising that crypto mining is added to the list.

Any and all heavy users of energy are contributing to climate change in some way, and the cost/benefit analysis for some are more justifiable than others. Food production is energy-hungry, but without it, people die. Nobody would die if crypto mining ended tomorrow. Thanks to government standards, idle electronics are better now than they were, and will continue to improve in the future. Big homes with A/C is a largely American problem without anything to really address it yet, but there has been a lot of attention focused on the issue already.

In general a bad thing is bad whether there are worse things or not. Crypto mining is a big drain on energy resources with very little benefit, or with benefit to very few people, and it doesn't have to the biggest drain, or benefit the fewest number of people the least for that to be true.

0. https://www.nrdc.org/stories/keep-your-devices-wasting-energ...

1. https://www.ajc.com/news/local/are-air-conditioners-slowly-k...

2. https://www.vox.com/the-highlight/2019/7/25/8881364/greta-th...

3. https://phys.org/news/2017-11-impact-climate-china-substanti...

By endlessly arguing about regulating minor consumer/production classes the attention span and opportunities for effective change diminish. The regulation should happen for the big polluters first, i.e. coal plants, transport ships. When that is done and the problem is largely fixed, you can spend your time arguing about cryptocurrency taxation, why not.

It makes no sense, in the current situation, to discuss 10000 small energy consumers and appropriate pollution tax for all of them. We need to fix the damn energy production first.

Human life in general is harmful to the planet.

So what are you saying?

I don't understand why you would feel as though this reply says anything. There is always something worse or has alternative consequences.

That it's important to measure the impact of each accurately so that we can make sensible decisions about where to curtail said damages.

Making the study about bitcoin in particular seems like a clickbaity way of getting attention, when the core issue seems to be dirty power generation. Bitcoin makes up a tiny fraction of non-essential American power usage, so focusing on it in particular just seems like jumping on the "Bitcoin is terrible for the environment" bandwagon. Articles on this topic seem to pop up every couple weeks on HN lately, and they generally aren't novel or interesting.

Are you saying we shouldn't understand the consequences of what we do?

It's useful to know that BTC has major externalities because it allows us to make a rational choice. We could make different choices about all the things you listed, too. Or not! But you seem to be actually arguing for willful blindness, here.

aRE yOu SayInG tHat

No Im actually not. You said that. Pointing out that you can cherry pick and micro-analyze anything.

It was a question. You seemed to be implying that any such study would be "fear-mongering" (your words). That suggests that you think the studies wouldn't be useful and that we shouldn't do them.

Your playground rhetoric isn't helpful.

You are misunderstanding the point. Of course, cryptocurrencies have externalities and it is good to quantify them. In that respect, the paper could be useful. The objection is, this paper makes no attempt to analyze and show cryptocurrencies are worth attention any more than other 1000 consumer classes. It just states "crypto has externalities", does not weight them properly in perspective and advocates for "more research" and "lets regulate". That is irresponsible. Regulation should happen at the biggest most important polluters such as transport and coal plants.

It's a single paper. It responsibly limits itself to a manageable conclusion and responsibly points out where there are still open questions.

Can you point out where in the paper the authors leap to "let's regulate" as you put it? There seems to be one small section about regulation, but it doesn't exactly advocate for it: it describes how their pricing research could be used in regulation. It describes challenges in regulating. But it responsibly avoids making moral/policy claims.

The whole paper shows implicit interest in taxing/regulating cryptocurrency mining instead of unbiased dispassionate scientific analysis. You can see this explicitly in the last section, starting with

> However, the ability to locate, and re-locate, cryptomining almost anywhere (e.g., following the cheapest, under-regulated electricity source) and fund mining camps from anywhere with complete or near anonymity create significant challenges to implementing effective regulation

Do you think a discussion of policy implications should somehow ignore relevant policy considerations?

It's helpful for authors to point out how they think their work fits into the broader scheme of things. They correctly identify that a high cost to benefit ratio might make people interested in regulating it; they also correctly identify that practical concerns make that difficult. There is no advocacy there.

Even if there were, it wouldn't be inappropriate to say "hey, there's a downside to this; we recommend trying to fix that."

People are playing with big enough money trading crypto that they throw in social media campaigns just for kicks.

They take the average emission per kWh generated in the country, combine that with existing research of the social cost of carbon done by the EPA (which includes damages done through climate change), as well as the change in mortality rate caused by these emissions. Then they combine this by the electricity requirements of bitcoin mining, and they have their results. (in very global terms, there's a lot more work done in the results).

I definitely wouldn't call it pseudo-science, as their arguments seem valid, and their methods seem to follow through at a cursory glance.

They do note some weaknesses in their research, such as not having access to the exact sources of electricity miners use, however, which might change the results.

Reply on news.ycombinator.com