That’s true. But it’s important to point out that proof-of-work (PoW) is more decentralized than e.g. proof-of-stake (PoS).
For example, with PoW you only need a single honest node to arrive at the correct chain during initial sync. With PoS there is no such guarantee, since there’s no inherent way for a new node joining the network to decide which chain to pick when presented with two different, but equally valid, chains.
>But it’s important to point out that proof-of-work (PoW) is more decentralized than e.g. proof-of-stake (PoS).
The opposite is true in multiple aspects. PoW has infinite economies of scale which inevitably leads to total centralization as all miners but the most efficient one have to shutdown.
PoS can stay decentralized forever because new stakers arriving aren't going to make existing staking unprofitable, forcing them to shutdown, as actual running costs are negligible.
Regarding resistance to government interference aspect PoS absolutely demolishes PoW. Overwhelming majority of mining is done in big industrial warehouses owned by registered companies. They're trivial to find, very hard to relocate and easy to take physical control of.
In contrast, staking only requires a node, which could conceivably even be on a satellite, on a ship, or anonymized using tor or other anonymity network.
>For example, with PoW you only need a single honest node to arrive at the correct chain during initial sync
This argument doesn't make sense, from the start the user has to choose some specific chain and has to download node software for it. Even soft forks starting from the same genesis block may require a manual choice, as it's possible for old nodes to diverge from the new if majority of miners ignores the soft fork.
Chain names are not defined by code over medium to long term, but by dynamic social consensus.
I've been mining for about 4 years now (1btc/month and 35-40eth/month since March 2017), and this is the first time I've heard someone trying to say PoS would lead to MORE decentralization. To be fair, I don't regularly converse online with other miners. I mean, I see what your arguments are here, but I'm just not sure I agree.
Usually the only people trying to promote PoS are the ones who are financially well-off enough to buy the masternodes. You could make the same argument for mining -- to do it on any worthwhile scale you better be prepared to spend a minimum of $500k -- but "worthwhile" is subjective, and some people are fine only making a residual income off of their mining, because they feel like they are contributing to the "cause" -- to keep the network flowing -- amongst other reasons. Initially, I got started with mining out of pure idealism about decentralization, anonymous currencies, and the cypherpunk roots of it all. That turned into me making significant investments to mine on a larger scale. Maybe people doing it for those reasons will be less important as time goes on and it's more mainstream...I don't know.
My other business is legal MJ cultivation in California and Michigan, and it's the same thing in that industry -- the only people wanting, or not minding, the expensive licensing and startup costs in California are the well-financed companies who can afford it all, who can wait many years for profits if necessary. Everyone else wants a way for easy entrance in the market, low licensing costs, etc.
As for government interference, I was a "traditional market" operator for ~14 years in the MJ business, and have experience with avoiding scrutiny from the government. In fact, the issues with bitcoin mining are identical to that of growing marijuana on a large scale -- exhausting large amounts of heat, the intake of large amounts of fresh air or cooling BTU's, using large amounts of electricity, the need for good airflow in the facility, etc. With that said, you could structure your mining operation so that only a very targeted, focused investigation could find you -- it wouldn't be easy to "fish" and find it. For example, the main indicator would be the electricity usage, but there are tons of legitimate businesses that use far more electricity than a mining operation. Establish a business that is in that industry, buy a warehouse for that purpose, get your electrical installed, and then run all your miners through your own VPN or Tor (I've never tried to run them on Tor, because I was always worried about speed/connection problems.) They would see large electrical usage, yes, but that is normal for your registered business.
In this hypothetical government overreach scenario, hiding a mining business -- if you really had to -- wouldn't be any more difficult than hiding a large scale grow operation, or a meth lab, or anything illicit really. You just have to know what you are doing, but people have been doing this for decades and decades without problem.
> this is the first time I've heard someone trying to say PoS would lead to MORE decentralization.
As time goes by, the number of people holding a token goes up, and competition among the staking services goes up, which results in more decentralization.
Take for instance on Tezos, with better internet we can eventually get RPi connected to Starlink which can be installed by the users themselves and start staking.
As hardware technology progresses, PoW goes in the opposite direction. Initially anyone can mine, but later only the most specialized optimized hardware will mine irrespective of the number of people holding the token.
>Establish a business that is in that industry, buy a warehouse for that purpose, get your electrical installed, and then run all your miners through your own VPN or Tor (I've never tried to run them on Tor, because I was always worried about speed/connection problems.) They would see large electrical usage, yes, but that is normal for your registered business.
The comparison with illegal mj grow operation breaks down because long-term profit margins on mining collapse to very small percentages, so risking prison for that would be idiotic, especially as legal operators (that eg. enforce some government blacklists) would have lower running costs due to not having to pretend they are something else. In addition, contrary to a mj grow op, you have to order asics from Taiwan/China at least once every few years which would be very hard to hide (gpu mining is going away).
This even assumes that you can order profitable asics at all, indefinitely in the future. The market is structured in such a way that really big mining companies are incentivized to make their own asics - the design of a sha256 asic is relatively trivial (compared to pretty much all other chips) and ordering directly from TSMC or Samsung would save them a fortune. For all we know, it's already happening - it's not like any mining op has an incentive to tell.
The deciding factor is that in PoW a majority can force anything on a minority - smart regulation would enforce blacklists via orphaning once compliant miners achieve a comfortable hashrate majority, totally killing the non-compliant operations. Legal miners would have an enormous financial incentive to comply: orphaning a minority of hashrate directly increases their revenue.
In the most extreme case, states could start mining themselves to take control of the industry - something that's borderline impossible in PoS, because they would have to buy the staking tokens from the market, and there's simply not enough supply available at any reasonable prices.
Then there's the defense issue: assume a PoW chain has been captured in such a way. If the PoW algorithm relies on asics, users could fork to a gpu PoW, but there's no second move - nothing can be done to fork away hostile gpu miners, short of going away from PoW entirely. Potentially, this could be used to kill _all_ PoW blockchains - a state first makes enough asics to obtain a majority, then kills the chain by exclusively mining empty blocks and orphaning everything else, then repeats the process with gpus. The victory would be absolute and the cost would be in the low XX billions of dollars.
Right now there's really no reason for any state to attack as non-speculative use is very limited - but imagine that Iran somehow switches to bitcoin for all external trade and relies on its bitcoin stash to function. Suddenly an attack becomes a very realistic possibility.
Even if somehow some state controlled enough stake in a PoS system to censor everything, the defense fork can be repeated indefinitely - by deleting deposits of a hostile majority staker(s) on a community fork. For this reason, I expect decentralized PoS networks (those that can be run on home machines and without on-chain governance, not cloud-run DPoS or PoS networks with governance) to never be attacked in such a way - there's no way to truly win, short of total global surveillance I guess.
A blockchain is, fundamentally, a consensus mechanism which decides which transactions are valid.
Proof of Stake is self-referential; the parties with value at stake have value because the blockchain says they do. It's a perpetual motion machine. When the shit hits the fan (say, a controversial fork) the people who thought they had control ("stake") will learn who really makes the decisions WRT which blockchain is canonical. That would be the exchanges, and what they agree on will rule no matter who stakes what.
The reason Proof of Work works is because someone has to pony up the compute power.
> Well of course they have to determine it themselves otherwise they are centralized and dependent on a third-party.
PoW doesn't have that problem.
> Please expand on what you think is flawed and what are the downsides.
If I make a parallel version of the chain where I got all the mining rewards from block one and ended up with a majority stake, how do you decide between that and the other chain?
>For example, with PoW you only need a single honest node to arrive at the correct chain during initial sync. With PoS there is no such guarantee, since there’s no inherent way for a new node joining the network to decide which chain to pick when presented with two different, but equally valid, chains.
For example, with PoW you only need a single honest node to arrive at the correct chain during initial sync. With PoS there is no such guarantee, since there’s no inherent way for a new node joining the network to decide which chain to pick when presented with two different, but equally valid, chains.