Bitcoin mining and energy consumption • Hacker News Board

Re: Bitcoin mining and energy consumption

#41

Post by seanwilson » Fri, Dec 08, 2017, 2:02 PM UTC

gtrubetskoy wrote: post #37

seanwilson wrote: post #22

One of the big selling points of Bitcoin is that it's meant to be decentralised but it doesn't look like this plays out. Proof of work seems to inevitably lead to all the power going to parties that can afford specialised hardware who are in countries with low electricity costs. Running a Bitcoin miner on commodity hardware now is pointless which seems to go against the spirit of Bitcoin when it started. Was this pre…

> Proof of stake is meant to consume less resources but then the power is then handed to people with the most money? I'm surprised nobody has caught on to the fact that the present fiat system is in essence a Proof-of-Stake one. > Coming up with a way to have a decentralised currency where everyone involved gets a fair say without consuming too many resources is a super interesting problem. Indeed. I almost wonder if…

> I'm surprised nobody has caught on to the fact that the present fiat system is in essence a Proof-of-Stake one.

That was what I was meaning really. If the people that have large amounts of money (whether that's in owning coins with proof of stake or owning hardware with proof of work/space) have a disproportionally large say in what happens it's not much of an improvement to the current systems we all live in.

Re: Bitcoin mining and energy consumption

#42

Post by gtrubetskoy » Fri, Dec 08, 2017, 2:06 PM UTC

ethagknight wrote: post #38

seanwilson wrote: post #22

One of the big selling points of Bitcoin is that it's meant to be decentralised but it doesn't look like this plays out. Proof of work seems to inevitably lead to all the power going to parties that can afford specialised hardware who are in countries with low electricity costs. Running a Bitcoin miner on commodity hardware now is pointless which seems to go against the spirit of Bitcoin when it started. Was this pre…

I remember reading that a key weakness to bitcoin was its susceptibility to fraud if a group of nefarious miners control a significant portion of compute power on chain, because verification is based on consensus. As such, a properly motivated bad actor can produce fraudulent results that can only be disproven by exorbitant compute resources. Obviously this would destroy the value of btc but they only need to keep it…

What we are learning is that there are far easier ways to steal bitcoin than the so-called 51% attack. Miner collusion would be so expensive to pull off and the amount of money you can steal this way is not that great - you still need the elliptic curve signatures to verify, so the only thing you can really do is "double spend" the bitcoin that you need to have in the first place.

Re: Bitcoin mining and energy consumption

#43

Post by jasode » Fri, Dec 08, 2017, 2:07 PM UTC

seanwilson wrote: post #22

One of the big selling points of Bitcoin is that it's meant to be decentralised but it doesn't look like this plays out. Proof of work seems to inevitably lead to all the power going to parties that can afford specialised hardware who are in countries with low electricity costs. Running a Bitcoin miner on commodity hardware now is pointless which seems to go against the spirit of Bitcoin when it started. Was this pre…

>Was this predicted at the inception of Bitcoin?

It was not predicted in the original Bitcoin whitepaper. Yes, it discussed the theoretical feasibility of a 50+% attack to double-spend but it didn't explicitly predict a "consolidation" to specialized hardware miners that leaves out the miners at home using regular computers.

>It's interesting how economics impacts the security of the protocol like this.

Every decentralized protocol suffers from unintended centralization caused by economics. The same thing happened to other protocols like NNTP (Usenet), SMTP, HTTP+HTML, and Git.

The underlying issue is that technical protocols still have to be realized on real hardware like cpus + harddrives + network bandwidth and consume real costs like human labor. You can decentralize a protocol specification but you can't decentralize the amount of money different entities are willing to spend on that protocol. Each protocol whether NNTP/SMTP/Git/bitcoin does not come with a $1 million grant for homeowners to spend at Newegg to keep protocols decentralized.

That's why protocols consolidate towards big players in a power law distribution.

Re: Bitcoin mining and energy consumption

#44

Post by Joeboy » Fri, Dec 08, 2017, 2:08 PM UTC

sandworm101 wrote: post #20

moneytide1 wrote: post #5

The electricity consumption is a red flag, but perhaps it could be cheaper than the cost to build and maintain Guarda/Loomis trucks that carry physical currency, energy to melt metal to mint coins, and all the costs of building brick&mortar banks (even then, bankers provide auxiliary services like loans and advisement). We are all spending time talking about it as a collective effort to determine whether or not it sh…

These numbers are far beyond the energy required for trucks hauling cash. And they dont fully account for the truck-type energy to build and manage the bcoin mining industry.

> truck-type energy

I think "embodied energy" is the technical term for that.

Re: Bitcoin mining and energy consumption

#45

Post by sova » Fri, Dec 08, 2017, 2:09 PM UTC

ethagknight wrote: post #38

seanwilson wrote: post #22

One of the big selling points of Bitcoin is that it's meant to be decentralised but it doesn't look like this plays out. Proof of work seems to inevitably lead to all the power going to parties that can afford specialised hardware who are in countries with low electricity costs. Running a Bitcoin miner on commodity hardware now is pointless which seems to go against the spirit of Bitcoin when it started. Was this pre…

I remember reading that a key weakness to bitcoin was its susceptibility to fraud if a group of nefarious miners control a significant portion of compute power on chain, because verification is based on consensus. As such, a properly motivated bad actor can produce fraudulent results that can only be disproven by exorbitant compute resources. Obviously this would destroy the value of btc but they only need to keep it…

Yes, the initial paragraph of the bitcoin paper (bitcoin.org/bitcoin.pdf) says that as long as 51% of the network is not attacking the network (that is, trying to forge their own, new, transaction history) the protocol is safe. Even when 51% of mining power is concentrated in the hands of a group, there is still a lot of reverse engineering to do to try and inject your own blocks as history. So far I have seen no demonstrable way to do this. If anyone is aware of research in this area I'd be interested to see how far it has come along.

Re: Bitcoin mining and energy consumption

#46

Post by sova » Fri, Dec 08, 2017, 2:10 PM UTC

albertgoeswoof wrote: post #25

sova wrote: post #21

Earlier quoted context omitted.

You are right, the comparison is based on a faulty premise: the premise that current money minting doesn't consume energy. You are very kind to point out that the pursuit of money is not an end-goal at all, and it would be great if a lot of these young bucks who made it big on bitcoin were investing their earnings into furthering the globe, instead of just living lavishly.

A lot of the bitcoin wealth is going straight into ICOs, which do free up capital for innovation, in theory

That is true! ICOs that have novel purposes are certainly welcome to the world. One day I'd like to make my own cryptocoin for my site so there can be a mini ecosystem / economy there, too.

Re: Bitcoin mining and energy consumption

#47

Post by mrb » Fri, Dec 08, 2017, 2:13 PM UTC

joss82 wrote: post #7

From a quick research online, it seems that most mining pools are located in China [1] Roughly 2/3rd of electricity in China is produced from burning coal or gas, which, upon combustion, emit their own weight in CO2, basically. [2] From this two facts, one can conclude that Bitcoin has a pretty huge carbon footprint overall. This is not the impression that I got from reading that article. [1] https://bit-media.org/bi…

You make the common mistake of assuming that if a pool is Chinese then its users are Chinese. In reality mining pools are used by people from all around the world. For example many like to mine at AntPool, because as the largest pool: their payouts are statistically frequent, they have a reputable and expansive infrastructure, eg. multiple stratum (API) endpoints in various countries which reduces latency for end-users.

As to miners actually physically located in China, many are located in China’s province of Sichuan specifically due to cheap hydropower, not coal.

Re: Bitcoin mining and energy consumption

#48

Post by seanwilson » Fri, Dec 08, 2017, 2:14 PM UTC

bkolobara wrote: post #36

seanwilson wrote: post #22

One of the big selling points of Bitcoin is that it's meant to be decentralised but it doesn't look like this plays out. Proof of work seems to inevitably lead to all the power going to parties that can afford specialised hardware who are in countries with low electricity costs. Running a Bitcoin miner on commodity hardware now is pointless which seems to go against the spirit of Bitcoin when it started. Was this pre…

There are a bunch of consensus protocols out there already. My favourite is the Stellar Consensus Protocol[1]. I think that modelling your consensus around entities you trust, instead of the most power or most most money, is more natural. [1]: https://www.stellar.org/papers/stellar-consensus-protocol.pd...

Awesome, thanks for the link. I'm not finding the paper that easy to read though...could you summarise how it compares to proof of work or proof of stake? The idea of electing which nodes you trust more than anonymous ones though does seem to mirror the real world though.

Re: Bitcoin mining and energy consumption

#49

Post by seanwilson » Fri, Dec 08, 2017, 2:20 PM UTC

jasode wrote: post #43

seanwilson wrote: post #22

One of the big selling points of Bitcoin is that it's meant to be decentralised but it doesn't look like this plays out. Proof of work seems to inevitably lead to all the power going to parties that can afford specialised hardware who are in countries with low electricity costs. Running a Bitcoin miner on commodity hardware now is pointless which seems to go against the spirit of Bitcoin when it started. Was this pre…

>Was this predicted at the inception of Bitcoin? It was not predicted in the original Bitcoin whitepaper. Yes, it discussed the theoretical feasibility of a 50+% attack to double-spend but it didn't explicitly predict a "consolidation" to specialized hardware miners that leaves out the miners at home using regular computers. >It's interesting how economics impacts the security of the protocol like this. Every decentr…

> It was not predicted in the original Bitcoin whitepaper.

> Every decentralized protocol suffers from unintended centralization caused by economics.

> That's why protocols consolidate towards big players in a power law distribution.

Hmm, given the amount of foresight in the original whitepaper I'm surprised the current situation wasn't predicted especially if it's common with other protocols that are intended to be decentralized.

Re: Bitcoin mining and energy consumption

#50

Post by phoyd » Fri, Dec 08, 2017, 2:23 PM UTC

digiconomist has a more "realistic" estimate by including real world factors like cooling costs, the operation of older, but still profitable hardware etc. into their balance sheet. Their number is 32 TWh over the year (vs 9,6 TWh in the article above)

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

Also, their yearly estimate was only 26 TWh a month ago. If we project this into the next year, then we'll end with 130 TWh at the end of 2018 - or roughly 40% of what the UK consumes.