Sort of, yeah. It would be 51% of verification nodes. If you bought 51% of all Bitcoin, but it was verified by everybody else in the network, you don't actually have any power to decide if your transactions were valid or not. If you controlled 51% of the nodes, you would have majority control over which transactions are valid.

it's not 51% of the nodes, it's 51% of the hashing power together with following the agreed upon rules of the network

You're correct. I oversimplified it quite a bit.

Bitcoin uses proof of work [1] to verify transactions on the network. For someone to effectively rewrite the network they would need to control 51% of the total computing power (in hashes/sec) on the bitcoin network.

[1] https://en.bitcoin.it/wiki/Proof_of_work

I see so I was mistaken about that. I still don’t understand why people don’t see that it would be trivial for a major bank (or now that Bitcoin has grown, a government) to look into making enough computational power they would destroy it.

For Bitcoin specifically (by far the largest), the amount of computational power is now very large indeed. An argument often seen against Bitcoin is that of it's environmental damages - quotes like "it uses more power than the whole of Denmark!".

So yes, 51% control could be usurped by a dedicated attacker - but the resources are no longer trivial. And as noted, this "control" doesn't buy you a great deal unless you keep it going for eternity, and for the time period you do have control all you can do is basically stop it working properly by preventing new transactions - rewriting history is exponentially more expensive the farther back you try to go.

Just having 51% of the network doesn't mean checkmate. It allows you to perform something called a double spend. I could buy something from you with bitcoin, while at the same time I attempted to send the same funds back to myself. If I controlled 51% of the network, there is a slim chance that would work, because the other 49% of the hash power is still validating the transaction where I sent you the funds, not the secret one I sent myself.

If this were a big transaction, I might choose to wait for more than one confirmation before completing my end of the transaction. The longer I wait, the more expensive it becomes for you to conduct your 51% attack.

It really is prohibitively expensive, to conduct such an attack, when you consider a few other factors. I could double spend the money to myself, but after news of such an attack came out, the price likely could crash, reducing the prize.

Not to mention opportunity cost. If you have the means to conduct a 51% attack it means you have a lot of capital and expertise. You could put those two together to make a giant pile of money through legitimate means, like operating a hedge fund. The risk is a lot less.

The only group that would be motivated enough and capable enough to perpetrate such an attack would be a state, and why would they bother when, like was mentioned elsewhere, they could attack bitcoin with the stroke of a pen, via regulation etc. It's waaay less effort.

As others have explained, it's arguable that banks cannot do this. Though I am curious if there some way to outlaw cryptocurrency, possible go as far as impose sanctions to other countries.

Also, would like to note that Bitcoin can also function as store of value just like Gold.

Any news about Lightning? Is that running live now?

Banks have decade old hardware and take forever to upgrade software and protocols. You expect them to design a sophisticated attack and run it for however long. On top of that, it would be really easy to find out which bank was doing it and that has huge legal and customer based ramifications. Definitely not worth it.

Lol, we a major consumer bank here that have "system error" and everyone's payday salary is mixed up. I imagine some miserable dev tasked to run a batch script and forgot to do some steps.

For sure these banks wouldn't use their own hardware, but hire a new team, new hardware etc.

Even for PoS cryptos where you require 51% of the currency to control the network, it would become prohibitively expensive to buy up 51% of the circulating supply due to market dynamics. PoS is arguably more secure than PoW. Arguably.

And then even if you do, you now own a lot of the currency and are in theory incentivized to keep the value from crashing.

Missiles have a lot of value. Crashing them expends that value. Governments still crash missiles into their enemies.

Let's go with a proof of stake coin like tron. They have a 2.5 billion market cap. That's more than all of the pentagons ai budget. Congress was bitching about the f35's 94 million dollar price tag. Do you think they are going to vote to give a budget for not only the attack but the person ell and hardware as well? All for a, according to HN apparently, "insignificant financial threat". Doubtful. Easier to just regulate it with a single pen stroke.

> My impression is if anyone has 51% of all bitcoin they can write whatever they’d like on the blockchain and validate it at will. Is this wrong?

Completely wrong. There is a set of rules which define whether a Bitcoin block is valid or not, things like "the block's hash has this many zero bits" (simplifying a bit here, the actual rule is a bit more complex), "all transactions have a valid signature" (simplifying a lot here, it actually runs a sort of a small program) and "no transaction spends outputs which have already been spent". It doesn't matter how much Bitcoin you have or how much hash power you have, even if you have 100% of all Bitcoin and 100% of all hash power, if the block doesn't follow the rules it will be ignored by all validating Bitcoin nodes.

What having 51% or more of the Bitcoin hash power gains you is only that you can rewrite history. You can present a chain of blocks to other nodes, and later present a different "longer" chain of blocks to the same nodes, and they'll accept the "longer" chain and discard the older one. All these blocks, however, still have to be valid to be accepted.