5 ms·

Humans are involved in banking transactions too, don't discount the carbon footprint of the collective of bankers and bank tellers. Those people won't go away

by yebyen 9y ago

Humans are involved in banking transactions too, don't discount the carbon footprint of the collective of bankers and bank tellers. Those people won't go away if we eliminated the whole banking system, but their carbon footprint will not have been wasted on the tracking and transfer of assets!

mikeash 9y ago
A single bitcoin transaction uses about as much energy as the average American uses in 10 hours. As long as those bankers are doing at least a couple of transactions per day, they still come out ahead.
- yebyen 9y ago
  "They get more done" =/= "They do more things" =/= "They are more efficient" There is downward pressure on the transaction volume of bitcoin for bitcoin users, because each individual transaction costs so much. This is not exactly an advantage of Bitcoin, I'll admit, but there are some people who will consider Bitcoin a success if ultimately it is only used for larger settlements between payment companies (they would prefer to see Bitcoin is not involved in micro-transactions, they would raise the tx fees to make this use case impracticable.) All I'm saying is, don't look at Bitcoin handling an average of 7 transactions per second and say "that's less efficient" because your average Bitcoin transaction is currently trending toward being significantly more complicated than your average credit card transaction. (And even after the recent improvements and pricing hype, I'm not actually betting on Bitcoin myself.) The BTC system, by design, is unsuitable for trustless person-to-person transactions in the form of Point-of-Sale. Every merchant who operates Bitcoin is trained to know that the transaction is not over until it's mined and in a block, and it's really much safer to wait at least a few blocks. So if you're selling coffee, maybe you don't worry about it because absolutely nobody is waiting 30 minutes for their cup of coffee, and if someone defrauds you more than twice you'll surely remember their face. By then you're still only out ~$10, and the cost of each attack is high, and likelihood of success is non-zero but fairly low. So you can eat the loss for the cup of coffee, sure, but this is definitely not a great sales pitch for Point-of-Sale system at BestBuy or basically anywhere else. Ethereum, on the other hand, mines a block on average every 14 seconds. You could realistically be safe and use that in a Point-of-Sale system. (And I'm not the expert, there may be even better systems than Ethereum. But don't say Visa... :-)
  - mikeash 9y ago
    The magnitude of the difference is so huge that I don't think it matters how complicated a bitcoin transaction is. It's something like a factor of 10,000 at the moment. And I think either way, my statement that the question is based on faulty assumptions is correct. If bitcoin supplants a large portion of the financial system then the energy use will be ridiculous. If bitcoin's energy use is kept low by limiting it to large settlements between payment companies, then it won't supplant a large portion of the financial system.
    - yebyen 9y ago
      I don't think that's a given. You're assuming that interest in mining will continue to grow at the same exponentially increasing rate forever, but that would most likely require the price to continue rising forever too. I don't think either of those scenarios sounds at all realistic. The bubble will eventually burst, but the blockchain will continue on. Bitcoin price can drop to $1000 tomorrow and many miners will lose interest on the following day, but the difficulty is what governs how much work is required to put a block onto the end of the chain; not the number of transactions. When the difficulty drops, (yeah it will bring some miners back). In this way, it is extremely beneficial that the price drops for the viability of the network. And sure, that is scary for speculators who may or may not understand it, but this isn't a frozen banana stand, it is a brand new monetary system. There is tons of risk involved for stakeholders. As solutions are implemented that make it easier to include more transactions in a single block, or to safely defer publishing the transactions into a block at all until some later date when there is less friction (SegWit does this?) the number of transactions that Bitcoin may process per second will rise, but the difficulty/hashrate are not a factor in this calculation. The price of bitcoin is what has driven the interest in mining, and the explosion of difficulty is what follows from that based on how the protocol is designed to work. These variables are not in any way dependent on transaction volume by number or complexity of transactions, though. You can absolutely increase or decrease the throughput of transactions handled by the network without necessarily having any dependent effect on energy usage, hashrate, or difficulty. (We could have a fine blockchain with only 5-10 people doing the mining, and it would almost certainly be able to handle just as many transactions, but there are quite a lot more people interested than that.)

by yebyen 9y ago

mikeash 9y ago

A single bitcoin transaction uses about as much energy as the average American uses in 10 hours. As long as those bankers are doing at least a couple of transactions per day, they still come out ahead.

yebyen 9y ago
"They get more done" =/= "They do more things" =/= "They are more efficient" There is downward pressure on the transaction volume of bitcoin for bitcoin users, because each individual transaction costs so much. This is not exactly an advantage of Bitcoin, I'll admit, but there are some people who will consider Bitcoin a success if ultimately it is only used for larger settlements between payment companies (they would prefer to see Bitcoin is not involved in micro-transactions, they would raise the tx fees to make this use case impracticable.) All I'm saying is, don't look at Bitcoin handling an average of 7 transactions per second and say "that's less efficient" because your average Bitcoin transaction is currently trending toward being significantly more complicated than your average credit card transaction. (And even after the recent improvements and pricing hype, I'm not actually betting on Bitcoin myself.) The BTC system, by design, is unsuitable for trustless person-to-person transactions in the form of Point-of-Sale. Every merchant who operates Bitcoin is trained to know that the transaction is not over until it's mined and in a block, and it's really much safer to wait at least a few blocks. So if you're selling coffee, maybe you don't worry about it because absolutely nobody is waiting 30 minutes for their cup of coffee, and if someone defrauds you more than twice you'll surely remember their face. By then you're still only out ~$10, and the cost of each attack is high, and likelihood of success is non-zero but fairly low. So you can eat the loss for the cup of coffee, sure, but this is definitely not a great sales pitch for Point-of-Sale system at BestBuy or basically anywhere else. Ethereum, on the other hand, mines a block on average every 14 seconds. You could realistically be safe and use that in a Point-of-Sale system. (And I'm not the expert, there may be even better systems than Ethereum. But don't say Visa... :-)
- mikeash 9y ago
  The magnitude of the difference is so huge that I don't think it matters how complicated a bitcoin transaction is. It's something like a factor of 10,000 at the moment. And I think either way, my statement that the question is based on faulty assumptions is correct. If bitcoin supplants a large portion of the financial system then the energy use will be ridiculous. If bitcoin's energy use is kept low by limiting it to large settlements between payment companies, then it won't supplant a large portion of the financial system.
  - yebyen 9y ago
    I don't think that's a given. You're assuming that interest in mining will continue to grow at the same exponentially increasing rate forever, but that would most likely require the price to continue rising forever too. I don't think either of those scenarios sounds at all realistic. The bubble will eventually burst, but the blockchain will continue on. Bitcoin price can drop to $1000 tomorrow and many miners will lose interest on the following day, but the difficulty is what governs how much work is required to put a block onto the end of the chain; not the number of transactions. When the difficulty drops, (yeah it will bring some miners back). In this way, it is extremely beneficial that the price drops for the viability of the network. And sure, that is scary for speculators who may or may not understand it, but this isn't a frozen banana stand, it is a brand new monetary system. There is tons of risk involved for stakeholders. As solutions are implemented that make it easier to include more transactions in a single block, or to safely defer publishing the transactions into a block at all until some later date when there is less friction (SegWit does this?) the number of transactions that Bitcoin may process per second will rise, but the difficulty/hashrate are not a factor in this calculation. The price of bitcoin is what has driven the interest in mining, and the explosion of difficulty is what follows from that based on how the protocol is designed to work. These variables are not in any way dependent on transaction volume by number or complexity of transactions, though. You can absolutely increase or decrease the throughput of transactions handled by the network without necessarily having any dependent effect on energy usage, hashrate, or difficulty. (We could have a fine blockchain with only 5-10 people doing the mining, and it would almost certainly be able to handle just as many transactions, but there are quite a lot more people interested than that.)