{"id":4654,"date":"2018-05-22T12:43:53","date_gmt":"2018-05-22T12:43:53","guid":{"rendered":"https:\/\/davidgerard.co.uk\/blockchain\/?p=4654"},"modified":"2019-04-26T16:05:48","modified_gmt":"2019-04-26T16:05:48","slug":"bitcoins-stupendous-power-waste-is-green-apparently-bad-excuses-for-proof-of-work","status":"publish","type":"post","link":"https:\/\/davidgerard.co.uk\/blockchain\/2018\/05\/22\/bitcoins-stupendous-power-waste-is-green-apparently-bad-excuses-for-proof-of-work\/","title":{"rendered":"Bitcoin’s stupendous power waste is green, apparently \u2014 bad excuses for Proof-of-Work"},"content":{"rendered":"

Food 200 Acres remaining arable land
\nData 100 zettabytes public ledger
\nBreathable Air 100 km3<\/sup>
\nDyson Sphere Cryptomining 380 yottawatts
\nLaborers 2 million
\nsomeone who is good at the economy please help me budget this. my planet is dying<\/p><\/blockquote>\n

\u2014 comedyblissoption<\/a> (after @dril<\/a>)<\/p>\n

 <\/p>\n

Bitcoin advocates will take literally any news that you might think was Bitcoin going badly, and say “this is actually good<\/i> news for Bitcoin!”<\/p>\n

Certainly all the news I post about Bitcoin is actually good<\/i> news for Bitcoin. There is no thing<\/i> that is not good news for Bitcoin. Exchanges kicked out of China?<\/a> Good news. Price crashes? Good news. Nuclear war sends civilisation spiraling down into the grim meathook Mad Max petrolpunk future? Good<\/i> news, probably.<\/p>\n

So … did you know that Bitcoin using 0.1% of all the electricity on earth, and 0.5% by the end of the year,<\/a> is good<\/i>?<\/p>\n

 <\/p>\n

\"\"<\/a><\/p>\n

Photo<\/a> by Hans at Pixabay, CC-0<\/small><\/p>\n

 <\/p>\n

Why Bitcoin mining and proof-of-work exist<\/h3>\n

Satoshi Nakamoto, Bitcoin’s inventor, needed a way to add verifiable transactions to a ledger, making sure nobody spent a coin twice\u00a0\u2014 but without any<\/em> central authority. This was the key requirement for Bitcoin.<\/p>\n

(As I go through in chapter two<\/a> of the book,<\/a> Bitcoin had political design requirements that don’t work and can’t work.<\/a> This is why it’s so weird and hard to understand, and why Bitcoin cultists shout twice as loud whenever it fails in yet another way.)<\/p>\n

Proof-of-work mining is an ugly kludge of a hack \u2014 but it was the only way Nakamoto could think of to do this.<\/p>\n

How Bitcoin mining works<\/h3>\n

Proof-of-work Bitcoin mining is built around wasting resources as fast as you possibly can.<\/p>\n

You may have read in the papers about Bitcoin mining involving “complex calculations.” This is incorrect \u2014 none of the calculations are complicated. It’s working out a hash<\/a> \u2014 a number calculated from a piece of data \u2014 like the check digit on your credit card number, but longer.<\/p>\n

It’s very quick to go from the data to the hash, but impossibly slow to start from a desired hash and guess the data that would generate it.<\/p>\n

A miner puts together a block of transactions that are waiting to be processed. They also tack on the hash of the previous block, and a random number (the “nonce”). They calculate the hash of the resulting block. If that hash is a small enough number … they win the bitcoins! And their block is added to the ledger \u2014 the blockchain.<\/a><\/p>\n

The calculations required to build the blockchain ledger could be done on a 2007 iPhone or a Raspberry Pi \u2014 all the rest of the electricity is literally wasted, just to run a lottery to decide who gets the bitcoins this time.<\/p>\n

All those computers doing Bitcoin mining just buy 14 sextillion lottery tickets every ten minutes, with one winner.<\/p>\n

That’s the “work” that mining wants “proof” of \u2014 generating lottery tickets. You show your commitment, and how much you deserve the Bitcoins, by wasting power faster than everyone else.<\/p>\n

If you want to win more, you need to add more computers, to generate more lottery tickets. This will soon lead to winners coming up more often than once every ten minutes. Nakamoto thought of this \u2014 and so, every 2016 blocks, the difficulty goes up or down, to keep it at one winner every ten minutes.<\/p>\n

Mining ends up in an evolutionary arms race \u2014 a Red Queen’s race, where you run as fast as you can to stay in the same place \u2014 leading to ever-increasing power usage, for about seven transactions a second maximum, and two to four in practice.<\/a><\/p>\n

Thus, Bitcoin is anti<\/i>-efficient \u2014 the more bitcoin is used, the fewer<\/i> transactions it can process per watt-hour. Bitcoin currently uses 300kWh per transaction,<\/a> and by the end of 2018 this will be 900kWh \u2014 for the same two to four transactions per second.<\/p>\n

Bitcoin uses as much electricity as all of Ireland.<\/em> And everyone else is starting to notice<\/a> \u2014 and they’re not happy.<\/p>\n

 <\/p>\n

\"\"<\/a><\/p>\n

Bitcoin hash rate,<\/a> January 2016 to January 2018<\/small><\/p>\n

Bitcoin gets less efficient with time, not more efficient<\/h3>\n

Non-technical people often assume that Bitcoin will get more efficient as it goes on \u2014 like other technologies do. This isn’t the case at all.<\/p>\n

With every other technology, the economic motivation is to reduce<\/em> energy costs. But with Bitcoin, you make your bitcoins by spending as absolutely much energy as you can throw at the problem.<\/em><\/p>\n

More efficient mining hardware comes out all the time \u2014 but it’s then set competing against other mining hardware of the same model, and the efficiency improvements don’t advance anything.<\/p>\n

Thus, Bitcoin’s energy efficiency only gets worse<\/em> with time \u2014 which is what we see.<\/p>\n

Hal Finney \u2014 the second Bitcoin user, and Satoshi Nakamoto’s beta tester for Bitcoin 0.1 \u2014 noticed a couple of weeks after Bitcoin 0.1 was released<\/a> that proof-of-work mining had obvious problems with externalities: “Thinking about how to reduce CO2 emissions from a widespread Bitcoin implementation.”<\/p>\n

 <\/p>\n

\n

Thinking about how to reduce CO2 emissions from a widespread Bitcoin implementation<\/p>\n

— halfin (@halfin) January 27, 2009<\/a><\/p><\/blockquote>\n