Topic: The Environmental Cost of Bitcoin - Youtube Video (Read 2769 times)

I agree it's not easy. The only viable solution so far is the mentioned "proof of stake". A migration to that would be quite painful, though, I guess. But if it's tested enough and the pressure to remove PoW is great enough, it could probably be done.

I fully agree with you except for "proof of work can easily be replaced." what do you mean by this? I don't think it's trivial to remove the PoW from bitcoin at all.

Don't worry, Facebook wastes a lot more power on complete trivia and drama every day.

Don't even get me started on Instagram or Twitter.

I watched the video, but it is all just bs really.
I think one have to look at mining as a way to distribute coins and more as a market function than "cost per transactions" and "kw per transactions".
When you strip away the incentive to get coins from mining, you could currently run the entire bitcoin on one low end PC, the database and transactions processing etc.
So when you don't get any more new coins the mining, then the cost of transactions becomes apparent.

Resources that go into mining ≊ reward. Simple as that.

The talk over-complicates things and yes I do see an environmental issue if Bitcoin replaces the Dollar as a reserve currency tomorrow but not when we have time to react to actual problems.

Proof of work provokes the reflex of "those that have are given more" but essentially that is not true. This reward could be essentially zero (yes, zero aka 0) as those with stakes in bitcoin don't need to get block reward in order to have an incentive to secure the network. Proof of stake "mining" could be an option in the client and as it would "only" cost anonymity, people would switch it on for some of their wallets.

Unfortunately Satoshi promised there will be 21million coins and getting them into circulation based on stake will not be acceptable for the majority of worker-class miners.

So instead of writing the same post I write every time the topic of Bitcoin energy usage comes up, I'll just link to all the others, in chronological order:

https://bitcointalksearch.org/topic/m.149581
https://bitcointalksearch.org/topic/m.156457
https://bitcointalksearch.org/topic/m.855455
https://bitcointalksearch.org/topic/m.1166103

And here's another one by Stephen Gornick, three weeks ago, pointing out the exact same thing:

https://bitcointalksearch.org/topic/m.1278020

Using the "4 MJ per banknote" figure in the presentation, and stats from the US Treasury, circulating one-dollar bills alone consumes 338 megawatts, or more than 50 times the entire Bitcoin network.  And that's even ignoring the fact (pointed out in the presentation but ignored in the figures) that the number of Bitcoin transactions can increase by a factor of 10 before hitting even the artificial transaction limits imposed by the protocol, which can be removed.  After that, you can read the wiki to see how easily Bitcoin can again scale by another factor of 300 with minimal increase in energy consumption.

Altogether, that gets Bitcoin to a market cap of around $10 trillion, while still beating paper money in energy consumption with basically zero modification at all.

Beyond that, if by some miracle Bitcoin ever comes remotely close to the stupid amount of waste generated by the traditional banking system (which it won't), proof of work can easily be replaced.

So energy usage is, again, probably the single biggest non-issue in Bitcoin ever.

I discussed in an other thread about how to use a more "usefull" algorithm for proof of work.

Folding protein would be a good candidate, but as I mentioned it's not easy to tweak it into doing what SHA-256 can do.

The key difference is that SHA-256 (or any other digest algorithm, for that matter) allows the miner to identify himself into the digest.  By doing so, it is not possible to "steel" the work of other.

Basically with digests the hash of the coinbase transaction is in the proof of work.   So its author can publish it and noone can use it to forge an other proof of work with an other coinbase transaction.  It's tough to imagine something similar with protein folding.

Yet I proposed something.

Let imagine we have a list of polypeptides whose we are interested in knowing their three-dimensional structure.  I guess this basically consists in knowing which conformation of their rotational angles (because from what I know there is a rotational degree of freedom for each peptid link, right?) gives the lowest energy.

My idea is that instead of working on each of these polypeptides brutally, as if there were all as important, each miner takes a preference list order, which would be the exact translation of the coinbase transaction.

So he tries to find a lowest geometric conformation for each polypetide (as compared to the previously known lowest ones), but with a amelioration factor (the ratio between the new lowest divided by the previous lowest) that would decrease in the exact very order of the list mentioned above.

By doing so, when publishing the result, it will be difficult if not impossible to re-use those results, since they are computed in a very specific performance order.  They will be usefull for scientists (they do indeed give a better known geometrical conformation), but useless for other miners willing to create an other proof of work for an other preference list order.

Yeah, I'm only about halfway through, but this and the other major assumption he makes, that somehow "creation" of Bitcoins is separate from transaction processing, means this analysis is pretty fundamentally flawed.

Also, I'm not sure I really agree with the way he just sort of guessed at the amount of "e-waste" required to support Bitcoin.  Though, I admit it is a valid concern that this is the first time I've seen anyone point out.

I like the reference made at the beginning, to the energy cost of a banknote.  It's kind of cool to think that having 25 banknotes costs you approximately 1 Watt.

Troll much?

Scratch an environmentalist, find a communist underneath. That's why they're called watermelons: green on the outside, red on the inside.

It is obvious the creator of that video has no understanding of anything beyond basic math. 

Obligatory XKCD:

http://xkcd.com/605/

But by how much? The US basically printed itself 1 trillion dollars for just the Iraq war or something (probably more). According to my calculations that is enough for at least 500 billion liters of gasoline.

Each liter is 36MJ according to my memory of wikipedia or 36MJ*1585/second = 57 GW averaged over the last 10 YEARS.

Of course that stuff is just the tip of the tip of the iceberg. Start to add in the real money spent on those wars (1 trillion per year not 10?), cars and mansions of banksters, the financial crisis and so on and you will puke.
I heard 40% of company earnings in the US were in the financial sector - which is then basically all waste.

You could definitely be for crypto currency on account of just wanting to "save the planet".

If I understand correctly, he says:

• energy cost to maintain EUR cash: 5 TWh
• energy cost to run Bitcoin currently: 0.043 TWh

hmmm.

why does he say it in such a weird way: "25 or 50 minutes of electrical power consumption of germany to produce 1 million coins".

he might then also say it takes 70 hours of electrical power consumption of germany to keep EUR cash floating around.

holy crap yes. that guy from the vid figures that "just to keep the EUR bills moving", 5 TWh are used every year (compare to energy consumption of germany of 600 TWh) that's friggin' significant, yes!

Except in this case the idea of Proof Of Stake (ppcoin) is relevant, because it claims to solve the problem of having to use proof of work in the long run.

I tend to agree that the video is flawed in many way's, to many assumptions and random variables that don't seem to fit.


He did not include them when talking about creation and "e-waste", he puts waist for bitcoin at 31-71 gram per btc and 1 gram per banknote (video 9:35). I guess including paper, ink, chemicals, metals, ... used in the production of a high grade bank note was a bit far fetched.

Lets just forget the fact that most money is created with a simple db entry ... but then again creation of bitcoin happens by verifying the transactions so one could claim that the creation is a simple db entry as well  

Hang on he did not include "e-waste" of fiat transactions.  

The main flaw in his reasoning comes in when he calculates the energy consumption of a single Bitcoin transaction (5...10 kWh), extrapolates that to all of Germany's electronic transactions (17 billion) and concludes that Bitcoin would need 85...170 TWh of electricity for Germany alone.
This logic is flawed because the energy cost is per block and never per transaction. In theory you could stuff a billion transactions into a single huge block and it would still use about the same energy to find the corresponding hash, as it does for a single transaction (at a given difficulty level).

The reasoning is about as sound as multiplying the value of all financial transactions in the EU by the weight of a single Euro coin and concluding that there's simply not enough metal in the world to run an economy on the Euro.

Also, the Bitcoin network currently greatly subsidizes mining by having a really high block reward - this makes Bitcoin transactions currently much more secure than would be justified by the amount of money transfered alone, but it also means a much higher energy consumption. This is however only a bootstrapping issue and with the block reward decreasing we'll have a much better energy footprint.

Finally, upcoming changes like for example transaction replacement makes it possible for most transactions to be settled outside the blockchain without sacrificing security and without requiring mega-blocks.

Environmental Protection: The New Socialism?

Has the inventor of the Internet, Al Gore, started his machine again?

short answer yes.

but even today bitcoin is ecologically better then fiat.