[buttershakes]

I think this is incredibly overblown. How much does the existing financial system cost to operate? How much do ATMs, Private financial networks, money security, transportation, staff, legal, etc, etc. Comparing per transaction cost does not even begin to account for the energy costs of the current financial system. Even with the escalating cost of mining we are getting a deal compared to the way things are now. Hopefully research will yield significantly better incentive structures than pure proof-of-work, but in the mean time it has proven to be remarkably effective.

[JumpCrisscross]

> Comparing per transaction cost does not even begin to account for the energy costs of the current financial system

Finance is about 7.3% of U.S. GDP [1]. The U.S. produces about 1,000 gigawatts of energy per year [2]. A first-order estimation thus yields 73 Gigawatts for finance. So about 150 times Bitcoin’s 500 megawatts.

About $700 million of BTC change hands every day [3]. That’s $250 billion a year. U.S. non-cash transaction volume is like $180 trillion a year [4]. So about 720 times the size of the Bitcoin economy.

Bitcoin thus appears about 5 times less efficient than the non-cash status quo. Given finance is less energy intensive than most of the U.S. economy, this is likely a lower-bound estimate.

[1] https://www.selectusa.gov/financial-services-industry-united...

[2] https://en.m.wikipedia.org/wiki/Electricity_sector_of_the_Un...

[3] https://www.quandl.com/data/BCHAIN/ETRVU-Bitcoin-Estimated-T...

[4] https://www.federalreserve.gov/newsevents/press/other/2016-p...

[whytaka]

I think the biggest cost to sustaining fiat is the power of the sovereign that backs it up. To a good extend, you have to also account for the military that secures the borders as that which also supports the currency.

[buttershakes]

Sure, but this doesn't account for how Bitcoin or digital currency actually scales. There isn't any reason that the same proof-of-work couldn't secure an order of magnitude more transaction volume. That kind of throws this analysis into question.

[acty1]

We talk about scalability in software where you can support 10x the value or transactions easily.

Bitcoin et al will easily support 10x the value and transactions and make the "but it is 5 times less efficient" argument moot.

Cars were also less efficient than horses at first. Maybe cars were a huge waste of energy and should have stuck with animal power.

[nnfy]

You simply cannot justify that first order approximation. There is no reason so believe that the financial industry takes power in proportion to it's relative size of GDP. In fact I would expect the financial industry to require a higher proportion of power to keep servers/ATMs/staff computers running.

[hwayne]

> How much does the existing financial system cost to operate? How much do ATMs, Private financial networks, money security, transportation, staff, legal, etc, etc. Comparing per transaction cost does not even begin to account for the energy costs of the current financial system.

The current financial system does things besides verify transactions.

- Need to get a loan, a line of credit, or just a safe third party to hold your bitcoin so that you're not carrying around your life savings in your pocket? Banks.

- Want to buy cats with your bitcoin but your cat dealer only accepts bitcoin cash? Exchanges.

- Want to do any kind of business in commodities? Now you've got derivatives, and you'll definitely want lawyers.

- Want to make any kind of investment? Now you've got everything.

The actual transferring of money only makes up a tiny fraction of the services in our financial system, and none of the rest of the stuff goes away just because you changed the type of money.

[menacingly]

But the point is those energy uses are incidental, as opposed to a cleverly-designed race of burn being intrinsic to mining, right?

I can see the argument that "waste" is the wrong word, since it is producing value, but it's worth putting thought into.

We as a race are really good at only considering the downsides of something once we've deployed it at irreversible scale

[nnfy]

>We as a race are really good at only considering the downsides of something once we've deployed it at irreversible scale

That's a little pessimistic, I think. Some problems simply do not become apparent and are difficult to anticipate prior to a certain minimal scale. This IMO isn't an issue with humans in general, our resources are limited in the design phase. There is only so much we can anticipate-and how often do you hear about large scale successes with the same magnitude of emotionally loaded reporting that you get with large scale disasters?

The fact that mining requires energy is not any different from literally any other action human beings perform to transact value. Bitcoin is just much closer to a pure exchange of information (which is all that value transfer really is) and so it's easier I think to stop and say "look at all those computers wasting energy doing nothing!" Since you typically see other "necessary" uses of power in non BTC transactions.

Not to mention the various solutions that could substantially improve BTC efficiency in the way of both hardware and software (e.g. asics). The fact that the BTC network is energy intensive is not as big of a hit against BTC overall as people seem to be making it out to be.

Not to mention that everyone is screaming about the energy use without considering both cost and benefit.