The bitcoin network already consumes an estimated 32.5 terawatt-hours per year. That's more energy than you'd need to power all of Houston. It's more than you'd need to power all of Ireland, and roughly as much as Denmark. Bitcoin miners in China suck lots of clean and cheap electric power from hydroelectric dams.
That raises the question: Is all this energy use worth it?
It's frankly hard to see how.
Bitcoin was supposed to be a kind of anarchist project: A peer-to-peer currency outside of the control of any government. There are certainly upsides to that premise. For people living in less-fortunate parts of the world, bitcoin can be a way for them to protect their assets or engage in economic activity out from under the heel of oppressive governments and failed states.
But that means bitcoin also has to get by on do-it-yourself projects by a decentralized network of enthusiasts. The digital process by which new bitcoins are created and added to the currency supply — called "mining" — involves a ton of computational power, and thus a ton of energy. Mining is also bound up with the creation and maintenance of blockchains — the digital ledgers that allow the community of bitcoin users to legitimate purchases and avoid fraud.
All of that complexity requires much, much more energy than established systems for payments and transactions. The Guardian reported that one of Visa's two American data centers uses about 2 percent of bitcoin's power consumption. Meanwhile, the two centers handle over 570 times as many transactions as the bitcoin network each day.
But it's also important to consider what people actually use bitcoins for.
There are certainly good and defensible ways to use a currency designed to evade government and law enforcement oversight. There are lots of bad and indefensible ways, too. (See: The Dread Pirate Roberts.)
But because bitcoin doesn't play nice with established governments and legal systems, and because of its extreme volatility and transaction fees, it's simply not a reliable way to actually purchase goods and services. The video game distribution platform Steam, for instance, recently stopped accepting bitcoins as payment.
What people mainly do with bitcoins is stockpile them and wait for their value to skyrocket. (I mean, wouldn't you want to invest in an instrument that went from $1,200 to $17,000 in less than a year?)
So we're basically burning those 32.5 terawatt-hours per year to power a particularly weird form of gambling. Which is doubly silly once you consider most of us in the developed world can already gamble with the money in our regular government-sanctioned checking accounts — either at a casino or on Wall Street. In raw economic terms, this pretty clearly doesn't qualify as productive activity or an efficient use of resources.
Bitcoin investing will inevitably remain a hobby for wealthy people in the developed world. Which undercuts the case that bitcoin-style cryptocurrencies help the developing world much.
Now, are there ways bitcoin could solve its energy problem?
Bitcoin is designed to make the computations easier as the supply of bitcoins grows. That should slow down its energy consumption per computation over time. Also, bitcoin's energy use is driven by the value of each bitcoin: As that price rises, the return on mining bitcoins keeps going up, driving miners to invest more in computational power and chew up more energy. Another big bust in bitcoin's price would lower its energy consumption.
Given the digital currency's meteoric rise, most analysts understandably expect it to plummet. But the last big pop in a bitcoin bubble was brought about by a big hack of one of the currency's key servers. That was a random event, so who knows if another precipitous fall is imminent.
Bitcoin could also just change.
Interestingly, it's designed to have a hard supply limit: Once miners create 21 million bitcoins, they can't make anymore. Again, that's thanks to bitcoin's crankish libertarian origins: Its creators wanted it to be a modern digital version of the gold standard. This is a terrible idea: It introduces all sorts of instabilities and trends towards deflationary recessions in the bitcoin economy. It also seems to introduce a lot more computational complexity into the mining and blockchain process than is strictly necessary. Ditching the 21 million limit could help with energy consumption, not to mention make bitcoin a more workable currency long-term.
Finally, if bitcoin's network and culture adapted to become more society and government-friendly, that ought to also open up opportunities to change its digital architecture and improve its energy use.
Of course, those two options would require bitcoin to abandon the spirit of bitcoin.
After looking at bitcoin's current trend, Wired concluded that "by February 2020, it will use as much electricity as the entire world." That obviously can't happen. And things that can't continue generally don't. The only question is, just how will it stop?
My guess is bitcoin will eventually correct itself by integrating with society, or another random disaster will knock the price back down. But I suppose it's also possible the digital currency eventually precipitates an international energy crisis, and that's how the unsustainable trend finally stops.
Obviously, it would be better to pump the brakes.