Hacker News Reader

Comment by michaelwww | Hacker News Reader

This is a great article. Buterin's critique of Bitcoin was very interesting.

Which critique? I searched for bitcoin in the article and didn’t find anything critiquing Bitcoin itself.

This bit in particular was interesting, talking about the comparative security of proof-of-stake Ethereum and proof-of-work Bitcoin.

> In the case of Bitcoin, I'm worried for two reasons. First, in the long term, Bitcoin security is going to come entirely from fees, and Bitcoin is just not succeeding at getting the level of fee revenue required to secure what could be a multi-trillion-dollar system. Bitcoin fees are about $300,000 per day and haven't really grown that much over the last five years. Ethereum is much more successful at this, because the Ethereum blockchain is much more designed to support usage and applications. Second, proof of work provides much less security per dollar spent on transaction fees than proof of stake, and Bitcoin migrating away from proof of work seems to be politically infeasible. What would a future look like when there's $5 trillion of Bitcoin, but it only takes $5 billion to attack the chain?

I can't bold here, but if I could I would have bolded "What would a future look like when there's $5 trillion of Bitcoin, but it only takes $5 billion to attack the chain? ".

Yes, that's a good one. He goes on to say"Of course, if Bitcoin actually gets attacked, I do expect that the political will to switch to at least hybrid proof of stake will quickly appear, but I expect that to be a painful transition."

which I thought was funny because he's tweaking them a little by saying they're eventually going to have to accept PoS anyway

If so, that won't be the Bitcoin my node will switch to. You can't force the network to do what you want. UASF and the many forks before, during, and after the fork wars have proven this.

Hey, you would be free to fork of to your own little chain that nobody uses. You can always do that.

But if the whole system is at risk due to attacks, I expect that you will be a bit lonely on that chain

I would just not do anything and continue on the original chain. Yes, if most of the world moves to the other chain, that fork will be Bitcoin for the rest of the world. The issue is if my time preference is low enough to await/speculate on the demise of the chain due to the issues I've identified and made me not switch.

Pretty sure node runners would prefer a UASF to hybrid PoS or at least a different PoW algo over constant 51% attacks.

I love Bitcoin, but you have to be able to consider hypotheticals like this. Similarly... a tail emission should be on the table. Mentioning it used to be totally third rail [0], but these days folks like Peter Todd are speaking up.

[0] - https://twitter.com/mhluongo/status/1092559017178734594?t=i2...

Consistent 51% attacks rely on Bitcoin already being more-or-less worthless. If it is, then yes, the cost of terrorizing the network will be low. The same follows based on Vitalik’s argument for why PoS is supposedly too expensive to attack.

The difference is if you want to terrorize Bitcoin, you need to find energy and compute at relatively low cost compared to everyone still on the network. With Ethereum, you just need to acquire ETH.

Attacking bitcoin uses energy but does not degrade the computer or the energy infrastructure required to do the attack. Attacking PoS ethereum degrades the resource you use to attack.

When bitcoin was created there weren't hashpower markets where you could attack a chain without ever having invested in it and then move on the next minute, your hardware investment still earning you money. I think satoshi thought there would only be one chain and the game theory with many is different.

Considering he thought Bitcoin block space was too scarce, and now not scarce enough, and considered people running their own nodes a “weird mountain man fantasy”, but has now built a political topology that depends on something like Bitcoin’s UASF and thus on people running their own nodes, I wouldn’t trust his predictions much.

The attack surface of Ethereum’s PoS is pretty huge, and derivative staking already has a runaway accumulation problem. Bitcoin’s entire critique of money is based on not being something like Proof of Stake.

Where could I educate myself on the vulnerabilities of PoS? It seems like for PoW, the vulnerability comes from variability of computing power of miners - if miners capacity dips for a few hours because of external events, attacker can use that opportunity and bring online a lot of resources to do a 51% attack.

Doesn't a PoS attack require 51% of coin ownership? How would an attacker obtain that without being massively rich to begin with?

A few are here: https://www.lynalden.com/proof-of-stake/

There’s also the overall economic argument that Proof of Work is actually just the least obscure version of anything else: https://www.truthcoin.info/blog/pow-cheapest/

Overall I’d just argue that because PoS doesn’t involve external resources to create blocks, the cost of producing a block won’t be well understood and thus there will be opportunities for rent extraction - which we’re already seeing with Lido staking.

Great article, thanks!

>>Considering he thought Bitcoin block space was too scarce, and now not scarce enough,

He never said that. He said there isn't enough fee revenue. Fee revenue can increase as blockchain space becomes less scarce, and more bytes are sold to transaction generators.

>>that depends on something like Bitcoin’s UASF and thus on people running their own nodes, I wouldn’t trust his predictions much

An Ethereum based UASF by no means requires every one running their own nodes. It just requires people recognizing - via either an upgrade to their own node, or choosing a third party node that has upgraded to pull data from - the chain that slashes the coins of those who are executing 51% attacks.

What would the absolute worst outcome of that attack be? Miners don't have the option of spending from addresses they don't have the private keys of; all they can do is rewrite history to a certain extent.

Would it be bothersome? Yes. Would a globally used bitcoin monetary base layer have complete backing of nation-states using it? Yes. Would people be hesitant to transact during the attack? Probably.

None of these things are of immediate consequence, while the attacker spends its funds rewriting history and getting blowback from everyone online.

Then there is the issue of getting to 51% in the first place, and retaining that 51% once it is clear something nefarious is happening. Many miningpools are conglomerates of individual miners, who certainly have the option of switching to other parties at their whim. I think proof of work is the only viable way currently known to man to ensure a fair, even monetary playing field at global scale. PoS is not proven, and Ethereum does not scale because it optimized for parameters that aren't relevant for a system that is supposed to be online for at least a century, if not longer.

> Then there is the issue of getting to 51% in the first place, and retaining that 51% once it is clear something nefarious is happening.

ETC currently has a 6.5 day deposit confirmation time on Kraken because it has had multiple 51% attacks and that hold is even after ETC implemented mitigations [1]. These are not just hypothetical issues.

[1] https://coingeek.com/ethereum-classic-implements-51-attack-d...

The reason this is possible for etc is because it's a small chain nobody uses with blocks that can be mined with hardware from much larger chains. There is no larger compatible proof of work chain than Bitcoin.

Correct. There can only be one top algo/coin per class of hardware.

Once ETH PoS happens, ETC will be the top GPU coin (at least to start) and the deposit times should go back down again.

One thing ETC has going for it is that the algo is the same as ETH. Very large farms will struggle to retune GPUs for different algo workloads. There is also the issue of different power requirements too.

This makes me wonder if I should speculate on the future price of ETC rising right after the merge, because most miners will switch their hashing power to that chain.

My theory is the price will jump up on the speculation and then head to zero as the large miners dump on dumb retail.

Just reads like a pitch?

Might be valid but Vitalik has kind of a lot skin in the game so it's hard for me to take it at face value.

I have the opposite attitude. I want to hear the best arguments from PoS people and the best PoW people. I'm not putting money into it until I understand it to the best of my ability, which admittedly may be never because it's so far over my head

There's a famous quote by Uptain Sinclair that is applicable, if indirectly. "It is difficult to get a man to understand something when his salary depends upon his not understanding it." If Vitalik was wrong, do you think he'd be able to see it?

That's a good point if I were forced to make a decision by only listening to Vitalik. But he doesn't strike me as a profiteering salesman trying to get me to sign the dotted line based on just what he says. It's a huge debate that will invent the future so no doubt Bitcoiners will engage in the debate and tell me why Vtalik is wrong about Bitcoin and Proof of Stake. You wouldn't complain to the judge in a courtroom that the lawyer for the other side is biased. That's essentially what some people in this thread are saying. Instead, you'd engage the opposing lawyers arguments and come up with better ones. I'm on the jury, trying to make up my mind.

The court compels the lawyers to tell the truth or face consequences, so at the very least, out right blatent lies will be called as such by the other lawyers or the judge. Certainly things will sneak by so you should be weary of whatever is said in court, too.

However this interview is PR, not a court.

I'll be the judge of that. But seriously, this interview is just one piece of evidence I've been collecting in my crypto folder. Like everyone else except for people who get into the team spirit and join one side or the other. I should just invest in both an forget about it, but that's not my nature.

I liked your metaphor, by the way. Thank you for engaging!

Thank you too. I used to love speech and debate in high school where it was about ideas, not emotions

> If Vitalik was wrong, do you think he'd be able to see it?

Vitalik is one of the few people that prioritize the technology over money. He's often been critical of Ethereum, its price, and previous choices he and the rest of the Ethereum development community has made.

Is it wrong, though? I’m constantly surprised how incurious Bitcoiners are about what will happen when the block reward subsidy pool runs out. I still haven’t heard a convincing story about what will happen to the network security a few halvings from now, besides “who cares, I will have sold by then”

Donno! But would you take Tim Cooks opinion on Android phones? How about Elons on NASA? Gonna need a less invested source.

Honestly I’d probably learn more from hearing Tim Cook talk about iPhone vs Android, or Elon Musk talk about NASA vs SpaceX, than I would from hearing from almost any impartial observer. I’m by no means a fanboy of either, but I respect the fact that both of them have spent a bunch of their life learning about this stuff and understand it at a level I don’t. Of course I’d treat it as a biased source of information, but I’d still treat it as informative.

How could you, a non-expert, discern what parts were bias or not? You might believe you'd learn more things, but how much of it would be spun? What is left out, glossed over, or exaggerated?

> But would you take Tim Cooks opinion on Android phones? How about Elons on NASA

I sure would. I'm always interested in what people have to say about their competition. When Steve Balmer said iPhones were a losing idea, did I believe him? No, I thought he was making a mistake, to put it politely. But it was instructive to let me know where Microsoft's mindset was at.

It is wrong. It is conflating two things: market cap and total vulnerable funds in a reorg. Doing a 51% attack isn't free range to steal any bitcoin, all you can do is reverse a recent transaction. And in practice it needs to be very recent (like within the last hour) to be pulled off.

Let's work out an actual attack: I send 1000 btc to an exchange, let it confirm, exchange for something else (Ethereum, let's say) and withdraw. Once my ethereum withdraw is seen on that network, I start 51% attacking the bitcoin chain and unwind the dozen or so blocks that have occurred since I sent the deposit. Maybe this attack costs me 500 btc, but in doing so I get my 1000 btc back. So I effectively bought that ethereum at half price by defrauding the exchange. It needs to be a lot of currency because the 500 btc price of the reorg (made up number!) is a fixed cost, independent of the value of the transaction I'm reversing.

The total market cap of bitcoin doesn't enter into this. The most value I can obtain is the lesser of (1) how much bitcoin I have, and (2) what limit the exchange places on deposits. Exchanges (or other trading partners) can put a hold or delay on high-value deposits, which completely mitigates this attack.

Now the expense of the attack is dependent on the mining revenue per block. So if the mining revenue per block is $300k or so, to use a number from the article, then exchanges and other users of bitcoin should require extra confirmations for transactions > $1MM or so.

I don't know about you, but I think it's perfectly acceptable to say that settling a transaction worth $500MM will require waiting for at least a day's worth of confirmations. And that's all that's required to prevent this attack from being profitable at scale.

From what I understand, the primary mode of 51% attack is controlling 51% of hash power, such as renting a lot of GPU for a short term attack. This gives you the ability to censor the network, rendering it near useless and probably crashing the value of BTC. If the attacker has shorted BTC it may be profitable to crash the price.

I agree that a 51% attack just to reverse a transaction is unlikely, but given how disruptive it would be (as a break in an otherwise ~perfect security track record), there could be other motives. For example, someone could take a short position in Bitcoin. Or a government could attack it for any number of reasons.

Bitcoin had been 51% attacked before.

But the point is the total market cap doesn’t factor into any of those possible attacks, which is what I said was incorrect about Vitalik’s statement.

> Bitcoin had been 51% attacked before

Are you thinking of Bitcoin Gold?

No, Bitcoin. Google GHash.io

GHash.io controlled 51% of hash power for a short time, no attack was carried out. So no, there wasn't a 51% attack on BTC.

Well there is this famous post by Greg Maxwell where fees almost reached the block reward amount (pull out "champaign"): https://lists.linuxfoundation.org/pipermail/bitcoin-dev/2017...

As long as there is demand for block space, a market will exist. Also, block size can be increased. Bitcoin will need a hard fork anyway to fix the year 2106 problem (https://bitcointalk.org/index.php?topic=760).

"Also, block size can be increased". Politically, no, it can't. That is the hill that "Bitcoin" has chosen to die on.

That's fairly literal. There are a few ways of fixing the core problem (e.g. continued issuance, increased blocksize, proof of stake) but bitcoin political ossification makes it look more likely the chain will die rather than make the necessary changes.

Other forks have done it.

They've already chosen, no?

Fees have come down significantly since the Maxwell post, and haven't returned to that level since. Hence the problem.

>>As long as there is demand for block space, a market will exist.

Of course a market exists. But the current demand will not be enough to secure the network once the subsidy sees a couple halvings.

>>Also, block size can be increased

Both a block size increase and raising the 21 million coin limit are unacceptable according to Bitcoin's current leadership.

It’s a market. It’ll adjust itself.

Thank god you can't bold.

I would have responded that the security of the original Bitcoin protocol will likely have been rendered void by quantum computers long before the full supply has been minted.

The time when the last satoshi is mined is irrelevant. The security will be in question as soon as the block subsidy drops below the level of tx fees, which will only take a few decades. In the past decade, no progress has been made on the size of numbers that can be factored by a quantum computer. From https://en.wikipedia.org/wiki/Shor%27s_algorithm:

> Also, in 2012, the factorization of 21 was achieved, setting the record for the largest integer factored with Shor's algorithm. In 2019 an attempt was made to factor the number 35 using Shor's algorithm on an IBM Q System One, but the algorithm failed because of accumulating errors.

Young technologies can progress exponentially (Moore's law, Wright's law,...) with the right factors in place, so be careful when extrapolating progress over the next decade using the last decade. How much progress in neural-network-based image recognition has been made in 2002-2012, and how much in 2012-2022?

If quantum computers will ever work, by that time bitcoin will migrate to new algo, that’s not a big issue.

In the same sense that moving to PoS is 'not a big issue'. The Bitcoin protocol, using PoW and quantum-vulnerable keys, is defined in the Whitepaper and can't move anywhere. People in the Bitcoin community can move to a new protocol and try to create an air of continuity (which will likely happen, multiple times, creating questions about who the 'real' successor is).

Reply on news.ycombinator.com