[anonporridge]

Vanishingly unlikely. PoS would effectively destroy bitcoin's immutability, which is a huge part of the value proposition. And more importantly, a huge part of it's core principles.

Coins like Ethereum that have already proven that immutability isn't a core principle (DAO fork) are fine moving to PoS.

Most coins don't need their own PoW though. It's likely that we'll end up with one PoW chain and everyone else can periodically checkpoint themselves into bitcoin's immutable blockchain.

[trompetenaccoun]

Ethereum was running on PoW when that happened. If anything you've just made an argument against PoW and miner centralization, where they can just get together and roll back the chain. The same would be possible in Bitcoin and while it hasn't happened yet there, we've had instances of large mining pools trying to censor transactions. A PoS system with sufficient decentralization would actually be more resilient against both these threats than a PoW one where mining is done by ASICs, which means large mining centers as opposed to individuals running an Ethereum node from their homes.

[anonporridge]

This is a reasonable criticism. Let me try to counter.

PoW doesn't protect you from a large enough cabal of power (in Eth's case, developers and whale bag holders) controlling the direction of the chain, or censoring entities they don't like from continuing to contribute.

What it does do is ensure that this large cabal can't erase history that happened. The only reason Ethereum was able to fork was because it was still young, they moved very fast, and there was enough external incentive from all parties to ignore the lost mining costs. And even with all that, Ethereum Classic still exists with all the PoW post DAO hack preserved.

On the other hand, PoS has no cost associated with writing the blockchain as far back as you want. If a majority cabal of stakers agree to rewrite history, they just do it. They can rewrite the chain from the dawn of proof of stake with whatever changes they want and it looks equally valid as the "true chain". No energy intensive hashes needed.

With PoS, you have to compromise a large enough number of stakers to mutate the chain. Very little cost after that.

With PoW, you not only have to compromise the miners, but you need to force them to burn tons of real world energy to catch up to the current block to rewrite the chain. Insanely expensive and effectively impossible the farther back in time you go.

[trompetenaccoun]

Fair point, you'd need more energy to do that in Bitcoin, although I want to stress that by no means could you simply roll back months of PoS Ethereum 2.0 history at the snap of a finger.

The DAO reset was very controversial and people still talk about it years later. Prominent Ethereum devs have repeatedly stated there would not be a second one. This isn't to say it couldn't happen, but it would be a big deal and even more contentious and would lead to even more fallout than the first time, should it ever happen again. So it's not like big Eth holders can just get together and roll back the state at will.

Granted, it's still a concern any way you look at it for certain applications where you want absolute certainty over long periods of time, for example for having an immutable archive of events for future generations. Wouldn't it be great if all that electricity were used to provide a truly reliable historical record for the first time in human history? Bitcoin could be used to preserve news articles, encyclopedias and other such data. For everything else PoS chains seem a lot more practical though. We have to see the pros and cons, for most applications it makes no sense to sacrifice all the advantages good PoS implications bring for absolute long-term immutability.

[anonporridge]

I do see a day when every database that promises immutability periodically checkpoints the root of a hash tree of every piece of data they have into the bitcoin blockchain.

Whether it's the Internet Archive, the Library of Congress, or your personal git repo, if there's some piece of data that you may want to be able to prove hasn't been altered, you check it into the blockchain.

[quickthrower2]

What is the use? You can’t recover the data from the hash, and the hash doesn’t prove the data is truthful.

[anonporridge]

Of course it doesn't prove truth, but it does ensure that the data hasn't been tampered with since it was published.

For example, some powerful entity might be incentivized to try to change the text of old news stories to better support their current agenda. If you can change our understanding of history, you can potentially manipulate the future. Records that include proof of immutability will go a long way to making this kind of power play orders of magnitude harder and more expensive.

[liquidify]

DAO fork has absolutely nothing to do with PoS. This is nonsense.

[anonporridge]

Actually it did.

> There’s another wrinkle: The Ethereum Foundation is on track to switch to a new consensus mechanism called “proof of stake”, and in that scenario, anyone who owns 14% of all ether will have tremendous control over the future blockchain.

https://www.coindesk.com/understanding-dao-hack-journalists

[space_rock]

DAO fork showed Ethereum to be centralised similar to a company therefore a database or PoS for the blockchain is a better fit than PoW

[liquidify]

This is nonsensical. PoS is a mechanism for assuring transactions can get added to the block chain securely. It has nothing to do with whether the controlling interests are centralized or not.

[anonporridge]

If you could take control of all the miners, how much would it cost to rewrite the chain?

If you could take control of all the stakers, how much would it cost to rewrite the chain?

Answer this honestly and you'll see why people trust PoW much more than PoS.

[space_rock]

PoS is only suitable for more centralised projects. For Bitcoin it would remove the properties that make the network unique