Hacker News Reader

Comment by noncoml | Hacker News Reader

All this cryptocurrenty craziness doesn't help either.

I don't understand how the HN crowd that is usually so pro-environment, is also happy with the tons of wasted energy from crytptocurrency mining.

View on HN

It's still young technology and the energy efficiency will improve with the lightning network or proof of stake or different concepts like filecoin.

The societal gain from these technologies is potentially very large and will hopefully realign interests in the direction of saving the environment rather than bailing out businesses.. But the most sane definition of money is actually energy imo

Also comparing the energy cost of printing physical money with mining cryptocurrencies is favorable to CC's iirc.

> Also comparing the energy cost of printing physical money with mining cryptocurrencies is favorable to CC's iirc.

Really? Last I checked, taking the averaged of the top 10 most efficient mining setups, BitCoin network was pulling 9GW+.

https://twitter.com/layoric/status/906087502414352384

Another model here showing 18 TWh annual usage, that's still an insane amount of power.

https://digiconomist.net/bitcoin-energy-consumption

Proof of Work (and BitCoin in particular) is so damaging, really hoping another currency takes over that doesn't use so much power for 4 transactions per second..

> Really? Last I checked, taking the averaged of the top 10 most efficient mining setups, BitCoin network was pulling 9GW+.

And where is the comparison vs physical money? You know, with trucks moving around every day, and the cost of insurance, special facilities, salaries, etc...

Except that cash is used in about a third of consumer transactions, and basically not at all otherwise. Everything else is chain of trust and essentially free, in terms of energy and every other resource.

Besides, a single truck can move millions of dollars, and I doubt reaching out to take my money and then making change contributes very much to the salary of a shop keeper. Probably far less than scanning my phone and waiting an hour for 3 confirmations. The only way a cash transaction would use the same power as a Bitcoin one would be if I had to personally drive the money 1000 miles to its destination in a Tesla.

Physical currency per bill needs to travel once and in big quantities in the same truck. Then it circulates on "residual human energy" from pocket to pocket for years.

There is an absolute initial cost, but then there is none. CC servers (and digital banking in general) needs a permanent upkeep of power.

I think the cost for paper money is neglible. Coins however is more interesting. Their face value tends to be less, yet weight is substantial compared to a paper bill.

Here I can see more an argument for cost.

This actually was very relevant going back a bit in history when we used silver and gold for money. You get a whole issue of material cost vs it's face value as well.

In case of heavy deflation, people might just melt pennies (copper) and sell them for other (ie. international) currency. To increase the person's actual buying power.

There's also a general argument to be made for copper and in the past gold for currency, as these materials hold an inherit value for their atomic makeup and it's applicable use which simply can't be justified/compared vs it's value as a human usable currency.

In general, both currencies have their values.

Physical: needs no power upkeep, nor situational power and data connection to work.

This is great for reliability.

A nice thing about digital currency though, is the power we get. Instant transfer over physical irrelevant distances is pretty nice.

Another nice thing is I feel more comfortable with just 20 bucks cash. If I wanted to go out in a world without digital transfers, I would carry a lot with me, for the drinks, the food, buy a situational ticket for some party you come across, money for a taxi/train.

In a world without digital currency, "hitting jackpot" for a robber is statistically absolutely higher, and thus it would potentially increase crime levels quite a bit.

I don't know of any data that isolates the energy cost of transport and storage of cash. Transactions are tricky, you could use the salary of the worker for the 5-15 seconds (or however long the handling/counting of cash takes) and equate that to cost of electricity.

Given a retail worker might cost a business $.03 per minute, you are looking at 1c per transaction. Now let's look at BTC. At 9GW when I looked at it, it was processing ~4.5, lets say 5, transactions per second. A kilo-watt-hour (kWh) of electricity will wholesale for 6c. If we are doing 5tx/s, that's 18000 in 1 hour. 9GWh to do 18000 transactions

(9,000,000,000) / (3600 * 5)

Gives us, 500 kWh per transaction, at 6c gives us $30 per transaction.

You might be able to argue subsequent cash handling might take a few minutes per transaction handles, but even so.

So given cash transactions save at least $20 per transaction, cost of transport and storage IMO would be negligable. These costs drop further with something like credit card transactions.

The two biggest problems I see with BitCoin and most PoW currencies (and I'm going to single out BitCoin) are;

1. BitCoin has a maximum transaction speed well below what is required for a world (or country/more localised) enconomy, 7 Tx/s doesn't cut it. VISA apparently does well over 1000+. 2. As mining gets harder, and as the network tries to reach the theoretical transaction speed limit, energy usage grows even more out of control.

9GW is such a massive amount of electricity at the moment. Even as solar + wind + storage becomes cheaper to build and capacities grow, it'll still be a huge amount of electricity and there usage by the BitCoin network will only grow. This is more electricity usage than the state of NSW in Australia _generates_ most of the time. That powers the residential and commercial electrical needs of a 1st world state with 7.5M people.

I think the more cheap power that is available to currencies like BTC, they'll keep using more and more with diminishing returns.

Right now a single transaction on the bitcoin network uses as much energy as powering 7 homes for a day.

There's infrastructure costs, but Visa's network doesn't need that much to process a single POST request

Do you have a source on that ? It seems impressive !

It got shared a bit further down the thread:

https://digiconomist.net/bitcoin-energy-consumption

As I keep pointing out, energy costs of "classical" money is considered an upkeep that's to be reduced. You can earn money by making the process more efficient. In cryptocurrencies, energy use is considered a feature - people are incentivized to waste more and more electricity.

Or put differently, it's replacing a system that's O(log n) with respect to energy usage with a system that's O(n²), for questionable benefits.

The whole crypto-currency scene is increasingly seeming like a huge Ponzi scheme. I really wonder at what point popular opinion will no longer support wasting electrical power on mining a digital currency.

When the music stops, and hope of striking it rich dies...people are greedy and shortsighted in the aggregate.

In aggregate over long time (multiple generations), not so much. Genetics optimized a global equation after all.

The problem is that we might lack enough time for this to work.

True, given time and the resources to enjoy that time, we’re not monsters.

The problem is that we might lack enough time for this to work.

Alas... barring that miracle at least, and probably a few more like it.

> I don't understand how the HN crowd that is usually so pro-environment, is also happy with the tons of wasted energy from crytptocurrency mining

Iirc that's one of the points of proof-of-stake. PoS is nearly analogous to PoW in its platform implications - potential for centralization due to resource costs and so on - but doesn't rely on burning through processing power to be secured.

Paul Sztorc argues that PoS will burn as much electricity as PoW [1]. Buterin answers here [2].

[1] http://www.truthcoin.info/blog/pow-cheapest/

[2] https://github.com/ethereum/wiki/wiki/Proof-of-Stake-FAQ#doe...

That's not an accurate characterization of his argument (in [1]) about PoW vs. PoS. It's not about which one will burn more or less electricity - his point is that each burns an equivalent amount of economic value, be that electricity, capitaltime, liquidity, etc. PoS tends to burn less energy but compensates this by burning exactly as many dollars worth of capitaltime and liquidity.

My disagreement with this analysis is twofold:

- The capital that PoS burns has been created by switching to PoS. It's measured in ETH, and it's that extra value that hn commentors are referring to when they snidely remark that PoS is a good way to deflate your currency.

- The negative externalities of power consumption for PoW are both murderous and civilization-threatening. PoS has basically zero negative externalities.

You derailed the entire point of the article by taking a tangent into a completely different, essentially irrrelevant, discussion.

“Bitcoin's electricity consumption as a percentage of the world's electricity consumption 0.09%“

How about Bill Gates and Terrapower trying to build a new generation of nuclear power in China?

https://qz.com/627113/bill-gates-says-china-is-the-best-plac...

TerraPower is interesting, especially since it recycles previously enriched uranium. On a side note, as far as the future of nuclear power goes, I'm more interested in LFTR designs and I think Flibe Energy is going to someone to watch here in the near future.

http://flibe-energy.com/

Totally OT, but have they cracked the corrosion issues?

The last I remember hearing about it was that there is active research in a new Hastelloy-N plus Titanium alloy combo that supposedly eliminates corrosion and neutron embrittlement.

http://moltensalt.org.s3-website-us-east-1.amazonaws.com/ref...

I like the notion of Bi-Pb eutectic designs myself, but LFTR is much more likely in the near term.

...or Thorcon Power. Bitcoin mining puts an interesting economic incentive on cheap energy production and chip efficiency. That and its implications always fascinated me.

Mining is a placeholder for the real resource which is useful computation. There will be an economy based on distributed electricity and distributed computation eventually

Filecoin is another proof of work that doesn't involve burning cpu cycles.

If the energy comes from a source where gathering and expending it doesn't have any adverse environmental effects then it's hard to say using it is "wasteful" - the only real problem comes if you assume using more power means more coal is burned and the like. Heavy coin mining tends to be done around the cheapest energy sources like hydro, and in the interim before burning fossil fuels is fully phased out I don't think the additional use from coin mining is going to be that big of a deal compared to people using inefficient air conditioner setups, heating their homes with electric too much, leaving all the lights and the TV on when they go out, or a bunch of other things that can add up to the additional power miners use.

https://digiconomist.net/bitcoin-energy-consumption

I beg to differ - currently just Bitcoin mining uses as much energy as country of Iceland - anything consuming energy on a scale of entire country (even small one yes) is environmentally significant.

“Bitcoin's electricity consumption as a percentage of the world's electricity consumption 0.09%“

By definition 0.09% is not significant. Eliminating it would not significantly impact climate change.

Excellent. Please send me a check for this insignificant amount, at current wholesale (business consumption) rates.

What does the amount of money have to do with anything? If someone found a way to reduce the $20 trillion US deficit by 0.09%, and someone said that it’s not significant, would it rate the snarky “well, then send me a check for that amount”?

Would you optimize a program by refactoring a function that used 0.09% of the execution time?

Yes, I often have done such refactoring. It's hard to get that last 20% of speed, much less last 5% without making a large number of small changes in code, usually after getting accumulated trace results (however presented.) .1% = one in one thousand. Keep stacking up those improvements and you do get somewhere, and often you'll easily find one hundred very small improvements in a day (10% altogether!) if it's green code, along with the bigger inefficiencies to fix. Not to mention that ignoring a presently-small inefficiency is predicting that future use won't swell that into a problem - which can be most unwise. Also, looking for tiny improvements gives you a good chance to stumble upon larger improvements you'd missed earlier. Optimize much?

You’re doing a lot of hand waving: “One hundred small improvements in a day” That’s a lot of work in an 8-10 hour day...

“if it's green code,..”

My point is that you don’t start by looking for 0.09% improvements when you’re profiling your code.

As for climate change, we are looking for cleaner energy options. The world’s energy use is going to increase greatly as billions more come out of poverty. China, for example, is going to increase nuclear energy to be almost on par with the US over the next decade:

https://www.nextbigfuture.com/2017/09/china-will-more-than-d...

Ya made me look - but no, you weren't saying starting. Nothing of the kind.

How much of the economy uses Bitcoin, though? That's the denominator to use.

this is it, we reached peak strawman

Do explain.

argument: 0.09 is insignificant for climate impact

counterpoint: 0.09 is a lot of money for one person

the rebuttal has zero to do with the point given and even less to do with the overall thread about mining being wasteful.

But the point, precisely, is that the amount is actually a lot of money for most countries, never mind a single person. That any attempt at any real economy can't ignore 1 in one thousand - especially if it might balloon 100x. Not sure you really read carefully.

Please do not move goalposts - I wrote "environmentally significant" not "significantly impact climate change".

I don't think our technology is so advanced that producing 18TWh of energy per year can be done without impact on environment.

You are the one moving the goalposts because the article is about climate change.

“Bill Gates has committed his fortune to moving the world beyond fossil fuels and mitigating climate change”

Back to my original comment: “You derailed the entire point of the article by taking a tangent into a completely different...”

If we move beyond fossil fuels and use 100% wind, solar, and nuclear energy, for example, does the electricity use for Bitcoin matter?

I think 0.09% is mind blowing!

Think about it, almost a thousandth of world's electricity consumption completely wasted on BTC alone.

Actually there is a great article by Nick Szabo on the social cost of organizing by third party that Bitcoin replaces - if you take the energy it costs to run governments/banks with everything they utilize in terms of energy cost then that's what Bitcoin tries to 'free up' also in terms of energy usage (and still comes out ahead)

Cheap energy sources like hydro are usually connected to a grid, so that electricity could be used to reduce fossil fuel demand elsewhere if it wasn’t being used for mining.

Reply on news.ycombinator.com