Why Proof of Work is not secure enough for Bitcoin
Вставка
- Опубліковано 30 чер 2024
- Video going over why Proof of Work is not secure enough in the long term for Bitcoin to be used as a replacement for gold.
I own Bitcoin and am not worried about Bitcoin being attacked. This weakness of proof of work doesn't mean it will be attacked. It means that it's not secure enough for sovereign entities to actually use it as a replacement for gold so limits Bitcoins use case of being digital gold
I also wrote an article on the topic if you want a bit more detail
altexplainer.substack.com/p/w...
00:00 - Intro
02:09 - PoW Weakness - Фільми й анімація
very good video, idk why there is so few views xD
I don't get how you do your calculation for the cost of the 51% attack. Do you not do a 51% attack on bitcoin by buying 51% of the ASIC mining hardware or how does it work?
Miners are mining for a profit so will only spend up to what they receive in block rewards on ASIC's and electricity.
The calculations are for the block rewards received over a year as that's the most miners will spend on mining.
What about the case like today where AntPool and Foundry control more than 51% of the network?
Could a court order to both cause a 51% attack and would that be enough to destroy Bitcoin?
People say “the miners would just leave those pools” but is that actually how it works?
Yeah miners could just leave in that case. The mining pool decides the block and all miners contribute hash power to creating the block. If they disagree with the blocks that the mining pool is producing, they can switch to another mining pool and contribute hash power to a different block.
The attack I described is incentivising miners to join mining pools that produce empty or highly censored blocks because the mining pools are able to give extra rewards due to a subsidy from a country like Russia/China.
"There is no way to reward honest miners more than attackers"
In the case of a selective censorship attack, this may be true. But this is certainly not true for an empty block attack. During an empty block attack, the honest miners earn transaction fees but the attackers don't.
If the attacker has less than 51% of the hash power they will mine blocks as normal and collect transaction fees in order to stay undetected. It's very unlikely someone will do an empty block attack with less than 51% of the hash power.
If they have over 51% and do an empty block attack, they will simply not reference non empty blocks. This would mean the honest miners will be creating blocks with lots of rewards but these blocks won't be a part of the longest chain as the attacker wouldn't reference them. If they aren't in the longest chain, they won't be final so the honest miners won't be receiving these fees.
What do you think of Proof of stake? Would that be less suseptible to this kind of attack?
Stakers are much less reliant on the block reward than miners so would be much more secure against this type of attack
What if, in addition to solving a puzzle, a validator *also* had to stake some money?
This way, even if Russia controlled 51% of the BTC hashpower, they'd have to stake BTC they may not be able to afford.
Which would take priority? If there's a fork and fork A has the majority of the hashpower but fork B has the majority of the stake which one wins?
If proof of stake always wins over proof of work in a fork, why not just fully switch to proof of stake?
Would love to see @Bitcoin_University respond to this
But you can pay higher transaction fee to incentivise independent miners. So whoever is censoring will lose compared to honest miners
There is no method to pay a higher transaction fee to an independent miner. Any fees available to independent miners will also be available to the attacker.
But if you want to censor some set of transaction you can’t receive tip from them
The attacker will have 51% of the hash rate and won't be referencing anyone else's blocks. This means the longest chain will be the attackers chain and it will only contain blocks from the attacker so any blocks created by the independent miners won't be in the longest chain so won't be a part of the ledger.
The independent miners won't be receiving any fees or block rewards as their blocks won't be a part of the longest chain
If fees of censored transactions is very high it might be profitable to start competing with attacker. For example if cost of producing same amount of hash power as attacker produces is 1 dollar. And you can get 2 dollars from fees. It will be profitable to “attack” the attacker. But yes, if it happens it will probably ruin bitcoin completely.
Yeah there are ways to temporarily defend against the attack but they require coordination from a large group of independent miners in a public setting that the attacker can see and adapt to.
Bitcoin can always take the nuclear option and hard fork away from using proof of work but then I'm not sure if it would still be considered bitcoin without proof of work
Cardano resolved this issue
couldn't you say the same for every possible consensus protocol?
Miners for proof of work are much more dependant on the mining reward than stakers are on the staking reward.
If Russia were to offer a subsidy for staking in Russia you may get a bunch of stakers move to Russia but then if Russia asks them to do something malicious or they will take away the subsidy, the stakers can just say no as they aren't reliant on the subsidy to be profitable.
Some staking protocols even have slashing so if the stakers were to perform a malicious attack for Russia, they would be slashed so would need to make sure the subsidy from Russia will be more than the amount they get slashed by.
Of course if you had unlimited money you could attack any consensus method but the amount you need to attack proof of work seems much less than the amount to attack proof of stake.