there are atleast 1,500,000 asics mining the blocks 24/7 right now. split off into hundreds of thousands of individuals homes and thousands of asic farms. pooled into dozens of pools
yep BILLIONS of dollars of hardware decentralised. and a millions+ dollars an hour in electric(spread across hundreds of thousands of electric bill payers) spent mining blocks

there is not one single place to "attack" in regards to re-organising recent blocks of bitcoins blockchain

to even attempt to amass enough hashpower to outpace the network someone needs to have control of ANOTHER 1,500,000+ asics

this then means they have 50%+ of the hash power. to compete in the race to achieve and remain ahead of finding blocks faster then the rest of the network

but even at a 1.5m vs 1.5m+ is a 50%+ chance of their block being first to be the next one added to the blockchain. and 10 minutes later. still 50% chance which this time they may not win.

even then.. its not as simple as just changing any/all blocks
ill explain
the blocks are linked together. they cant just decide to change one block from year X in the middle of the blockchain.. they have to work backwards from the current block if they want to change older blocks

but before i get too far into that. lets just explain what they "could" do, which is 3 things:
1. make new blocks do "empty block" meaning not include transactions.
2. only include their prefered transactions in new blocks
3. go back only a certain number of blocks

lets delve into (3)
to go back 1 block and change its transaction list. requires then to re-mine that block. which takes time. then they have to mine the current block the clean network is mining and then the next block before the clean network does to outpace the clean network to then changed block list of the 3 blocks(past present and future) then become listed as the blocks the network follow.. but even this is not a 100% guarantee its a 51% guarantee they can win the race.

to go back 2 blocks takes twice the amount of time meaning even less change.

the costs of going back and editing a blocks transaction list costs more the fursther back you go with less and less success unless you have ALOT more then 51% of the clean networks asics

the cost of the attack becomes too high to even be profitable. becasue all you will achieve at best is to change out a large transaction you already spent. to make it appear that it was not spent. so you can spend it again..

but here is the thing. for any transaction of reasonable value. people do not accept it as immutible unless it has 6confirms.. thus knowing its safe to assume its immutable at that point

yep.
it costs over $300k of just electric (and billions of hardware) to re-org 1 block
double that for 2 blocks($600k) double that again for 3 blocks($1.2m)  and so on

so people that accept payments of lets say $50k know its not worth a malicious person to try to undo a $50k tx. so people accepting payments take the extra precaution of waiting for 6 confirms before assuming the transaction is immutible

yes they can assume immutibility of $50k at 1 confirm. but when $50k is involved, no point being impatient. just wait 10-20-50 more minutes and be rest assured its absolutely not worth a malicious person to try to double spend it.

oh and also any merchant doing sales/accepting payments of large amounts is also watching the network for pools that are suddenly using higher hashrates compared to other pools..

there are many many safeguards and watch to watch for possible issues and also just to take precautions if dealing with large amounts..

..
that said,,
many many altcoins with very low PoW hashrates can easily be chain re-org (51% attacked) very cheap and without much effort.. those altcoins only protection is that no one cares/uses those alts enough to even be worth attacking. thus although they can be easily attacked. "whats the point" mind set is their reason no one bothers

SOURCE
[/quote]

That's not what happens. 51% attack can undo the last blocks and replace them with new blocks, but the blocks still have to be valid. They can't edit the blockchain and transfer someone's bitcoins. Your coins can only be stolen if you received them recently. The more confirmations your coins have, the more expensive it is to steal them. It's not about just owning 51% of power, it's about paying for it to execute the attack. So if you receive $10,000,000 worth of BTC, it's better to wait for 50 confirmations or more, before viewing transaction as final. Although there has never been a documented case of a 51% attack on Bitcoin. But it happened with many altcoins.

If a node is compromised only the information on that node is false, other nodes can still be productive and can prove they have the longest chain.

A 51% attack would be quite impossible. If you wanted to do it now and you weren't a distributed attacker, you'd get rioters shutting you down because you've taken their power away from you. If you were a distributed attacker, you've got a hard job ahead of you for making the reward you're about to get worthless for no reason.

51% attacks are unlikely to replace transactions in blocks that have already been mined too - especially if they happened more than a few blocks. It becomes exponentially harder to remove a block the further back you go.

I don't think mentioning it as hack is correct term to use.

It's 51% attack or Replay attack. The cost is huge because total hashrate of Bitcoin network now is very huge, not like years ago. But even years ago, it did not happen. Do you think the Bitcoin network hashrate will continue to rise or will be decreased in future. If it rises more and more, the 51% attack becomes more impossible.

In the past, some big pools with correlated owners had more than 51% hashrate of the network but they did not make such attack. They are smart to know such attack will not give any benefit to their pools because Bitcoin will lose faith in investors, then it will collapse to the ground. No pool, no whale want that to happen.

The first part regarding 51% attack depends on what cryptocurrency we are talking about. Unfortunately when the quote says "blockchain" it can be any blockchain, for example if I create a PoW coin today with nothing to offer you can 51% attack it with your own CPU without really spending any money! But that doesn't work in bitcoin due to how huge the hashrate is and how well distributed the miners are.

The second part doesn't make any sense. If a single node is disabled, then that single node is disabled not the whole network and it doesn't affect the "blockchain" at all. After that single node comes back online it catches up to the rest of the decentralized network by connecting to multiple other peers. Which means there is still no "compromised" state defined for that disconnected node.

There was a time when we were close to a 51% attack... and you know what happened?

1. Some of the miners in the Ghash.io consortium/Pool .... left the Pool on their own account, because they knew a successful 51% attack would be bad for everyone.

2. Imagine that you managed to pull of a 51% attack and suddenly the media gets hold of that information.... well, the Bitcoin price will take a massive hit and it will go down like crazy... because it will show that the Blockchain security are not strong enough to prevent such attacks.

3. The attacker will also lose a shitload of money... because 51% of the total hashing power is incredibly expensive. 

Not exactly an assumption but, if you went through the information on source link I provided, you would see that the article concerns "decentralized internet" and when the topic of decentralization is brought forth the first pointer as one of the successful model of decentralization is cryptocurrency and when blockchain is mentioned in that regards, what that comes to mind is that of bitcoin. I believe if reference was made to another Blockchain, it would have been mentioned. Also you would notice that the quoted sentence is a harvest from Medium, I tried to see if I could get to the source article on Medium but I couldn't so I had to generally ask here for people's thoughts.

Technically speaking, yes. But I guess "hack" related issues, as far as I can remember, are more often involving user neglignecne, or simply involving human interaction such as, 'accidentally' sharing private keys, or logging to other devices in some instances, and encountering phishing sites. Not trying to sound cliche' but we must really take extra cautions with it, given that it involves our money in the first place. I have also heard some stories wherein 'hackers' managed to access their assets but I doubt they did nothing for those bad people to get access or to reach them, chabces are high that there were some errors, 'coz if there are no such thing, why are whales or big investors are not having problems of this kind then? Practically, they should be more of the target, right?

The 51% attack and you mentioned is about replacing an older blocks with another valid block meaning they both follow the same "correct rule" since they are both valid. You can't reject such blocks. Also technically it is not the miner rejecting invalid blocks, it is the node.
The reason why 51% attack doesn't work is because how decentralized the mining in bitcoin is. It is not possible to gain control of that much hashrate.