In Defense of Proof of Work
neptune.cash> The significant amount of electricity consumed by the mining process does not in itself emit CO2.
Yes it does.
[edit] Not only in the sense that you have to build the solar panels, build the wind farms, build the mining rigs, then throw out 97% of the mining rigs (!!) without them ever having validated a single block. [1]
But also in the sense that any additional proof-of-waste operations run on the side steer that renewable energy generation away from productive applications on the grid -- failing to help decarbonize the grid, something we desperately need to do.
[1] https://mobile.twitter.com/DigiEconomist/status/143889647759...
Negative electricity prices is an increasingly common phenomenon due to an increasing supply from electricity sources whose output can't be controlled, i.e., wind and solar. Proof of work mining sets a floor under electricity prices, as mining can be spun up when prices are low. This means that proof of work acts as a subsidy for wind and solar, and increases the proportion of the total power production that can be delivered from these sources.
Almost surreal reading this. The article suggests that energy consumption does not need to be considered because hydro-electicity exists. It's a fantasy that ignores any sense of economy, science and politics.
> It's a fantasy that ignores any sense of economy, science and politics
Welcome to the world of crypto!
Even if all PoW energy came from renewable sources, it's extremely wasteful and I'm sure we could have used that energy on something more productive.
That's not to say the allocation mechanism in PoW isn't interesting, but that's not really good enough in itself.
I've been experimenting with energy limited PoW, something called NuPoW [1], that reduces the reward if too much energy is collectively used. But this only works if you outsource the security to an underlying chain.
The ever more prevalent phenomenon of negative electricity prices is a market signal that the energy could not be used more productively. If it could, prices would never be negative.
While environmental damage is a serious downside to PoW, the alternative of PoS has some serious downsides as well:
1) It's not a fair coin distribution mechanism. A coin that launches with PoS has the creators holding the entire supply.
2) It's not objective. Resolving long range attacks requires social consensus on what is the correct chain (https://academy.binance.com/en/glossary/weak-subjectivity)
3) It's much more complex.
I believe that the downside of PoW can be greatly reduced by having the emission be purely linear (fixing the block subsidy). Taking 20 years to reduce the yearly supply inflation to 5% will strongly discourage speculation and keep the price low, in turn limiting miner energy use. It would encourage use as an actual means of exchange / currency.
I would argue that proof-of-work and proof-of-stake converge to the same thing, at the limit.
In the proof-of-work model, a few folks make a lot of money (either by being in the right forum in 2009, rugging a community or lord forbid by actually doing something productive for society). These folks then go and buy shares in a mining pool using the PoW coin. These shares are basically the same as staked coins. They can go back and un-stake them at any time by selling. As owners of the pool, their largesse grows as transactions are validating and block reward accrues conceptually to them.
This is conceptually identical to a proof-of-stake system where a few folks make a lot of money (either by being in the right forum in 2013, rugging a community, participating in a presale, or lord forbid, doing something useful for society). These folks then stake their tokens. As stakers, their largesse grows as transactions are validated and staking reward accrues to them.
> In the proof-of-work model, a few folks make a lot of money
That's more a consequence of the poor emission curves that were chosen in the past. Practically no-one has made money on a linear emission PoW coin.
> These folks then go and buy shares in a mining pool using the PoW coin.
The mining pool sharing business is fraught with scams. Most bona fide pools require miners to operate their own rigs.
If it's a currency they can earn it in a variety of ways. Money has intrinsic economy of scale and pools all by itself.
Good point with the subjectivity of PoS. But regarding the linear emissions proposal, do you have any source that addresses the economic consequences of this model?
Within 50 years of linear emission, the supply inflation will be what fiat is targetting (2%). It will keep lowering after that and behave more like gold.
I get the idea, I was just wondering if someone had written something about the projected wider societal impact of such a scheme.
But maybe that misses the point: I have a firm belief that crypto currencies in one form or another is the future of money. History is too fraught with hyperinflation for it to not happen. The crypto currency that wins, though, will be the one that's in highest demand, not the one that's serves some abstract societal goals. So maybe the relevant question is: would linear emission be more attractive for its holders than e.g. Bitcoin's emission scheme with halvings of the inflation rate every four years?
I don't think crypto is the future of money. But I do see linear emission to be more attractive for all future generations, as it maximizes the emission in their lifetime, and doesn't make them feel like fighting for breadcrumbs from the earlier generations.
I see some issues.
We can discard Hydro pretty much immediately. There's not a lot of it that's sitting unused, there's not much opportunity for growth.
If there was this much wasted energy, the problem would be solved already. If the price went negative really often, that'd be a perfect business model for building some storage. Storage on the whole should be simpler than a mining farm. Putting extra megawatts into a pump or electrolysis is a relatively simple proposition. The computer hardware needed to burn that power on demand is far more expensive and complex.
Mining can cross borders, but cryptocurrency usage much less so. Yes, you can mine all you want across the border, but as governments sour on crypto they're going to make it much harder to transact in it. And your crypto isn't good for much if you can't sell it and get $ or Euros in your bank account.
The countries most amenable to this kind of thing are not all that profitable. They've got weak economies, bad governments and bad infrastructure. They're also easily pressured by wealthier countries, so they can't do whatever they want.
Long term if this worked it'd lead to centralization. Once you find some optimum -- which is probably a company that produces solar panels around the equator, they win the game and are hard to out-compete.
And of course, in the end this is still all a pointless waste. It's still countless video cards, solar panels, ASICs, and other hardware being piled up higher and higher to just keep up a global competition going that doesn't really achieve any improvement. It doesn't add capacity, it doesn't improve speed, it doesn't solve any practical problems. It's just an arms race where you need an extra tank because the other side made another one.
Energy storage is the deus ex machina solution that's always thrown around when real engineering concerns about the problem of electricity sources whose output you cannot control are presented. I might as well flip your argument on its head: if it was as easy to store energy as you propose, then electricity prices would never become negative.
Government is an abstraction that might make us blind to the fact that political decisions are made and enforced by actual people. The more economic power that crypto currencies amass, the bigger a threat they become to big government's taxing abilities, yes. Granted. But at the same time it also affords its supporters more lobbying power. I think governments face a prisoners dilemma in this regard: most governments might be better off if they all banned crypto currencies. But even better off if only their neighbors banned them and they allowed them, so they would get an influx of people and businesses making money in this sector, and they could tax these individuals. Think e.g. Crypto Valley in Switzerland, or Bitcoin enthusiasts visiting (or moving to?) El Salvador. And we all know how the Prisoner's Dilemma plays out in the long run: to the detriment of its actors.
And to you last points: ask modern day Venezuelans or North Koreans whether it's a waste to have sound money that can't be inflated by the whim of self-serving autocrats. Or ask a German citizen that was alive 100 years ago if sound money is a waste of resources.
> I might as well flip your argument on its head: if it was as easy to store energy as you propose, then electricity prices would never become negative.
Then you're not understanding what I'm saying. I'm saying that cost/complexity for dealing with energy overproduction goes:
Ignore it -> Adjust existing systems -> Store it -> Use it for mining.
The fact that storage is still a small niche shows that there's not enough waste to make even simple storage clearly profitable, so we're still mostly on the "adjust" step.
If it was a big untapped market, we'd never get to the mining part, because pumping water uphill is far technologically simpler than setting up a data center full of video cards.
> Think e.g. Crypto Valley in Switzerland, or Bitcoin enthusiasts visiting (or moving to?) El Salvador. And we all know how the Prisoner's Dilemma plays out in the long run: to the detriment of its actors.
El Salvador is an absolutely nonsensical usage of BTC. Bitcoin has long dropped any pretenses of being a currency, and is a horrible fit for normal people. BTW, the experiment in El Salvador is not going well.
> And to you last points: ask modern day Venezuelans or North Koreans whether it's a waste to have sound money that can't be inflated by the whim of self-serving autocrats.
As far as I can gather, the experiment in Venezuela is failure thus far, and North Korea is completely a no-go, since the general population has no internet access and NK barely has electricity.
> Ignore it -> Adjust existing systems -> Store it -> Use it for mining
That's just, like, your opinion, man.
Sure, but it makes sense. You have overproduction, how do you deal with it?
If a tiny amount, you can ignore it. Extra boxes pile up a bit higher than ideal. Power production gets a bit uncomfortably close to the maximum limits established, but you don't get into any trouble. You take note, but it's not worth really doing anything yet.
A bit more, and you start making minor changes. Slow things down a bit. Offer discounts to get rid of excess stock. Talk to the nearby businesses to adjust their schedules to use more power when you have too much.
If you have a lot of excess but have good reasons to believe that it won't be piling up indefinitely, but that you will profit from buffering the excess and then selling it off in times of high demand, that's when you build storage.
And something like crypto mining would be my very last choice. You're talking about building highly complex, highly specialized infrastructure for a single use. Storage can sell to anyone who wants, while crypto hardware only does crypto. Should conditions change it'll be far less flexible than storage.
Saito consensus is simply superior on the cases author identifies. You can still use hashing to ensure fair payments without spending 100 percent of the block reward on it.
Is the binding of the payment to the act of producing blocks which is the real problem.
In the end, a lot of the value of PoW cryptocurrency comes from the burning of energy it takes, let's get last year's 110 TWh for Bitcoin, at US$ 0,10/kWh we get US$ 10b burnt in electricity just to sustain PoW...
Many share the conviction that proof of work mining is a relic of the past, that the future of blockchain lies with its competitor, proof of stake. I attempt to rebuke some of the criticism that proof of work faces and argue why it is here to stay.
They're both bad, and they both have their own problems.
You know what solves it? Trust and a legal system. Time to throw this failed experiment into the dumpster and move on.
The current US inflation rate is 8.5 %. In Argentina its 55.1 %. Since the founding of the Federal Reserve in 1913, the dollar has lost 97-98 % of its value. How's that trust working out for you?