[ninepoints]

> I think it's interesting that via Ethereum, I can spend Ether to perform distributed computations on nodes across the globe and that this also serves as a store of value.

I find this "value proposition" highly dubious. You can already rent compute globally and perform computation in an environment orders of magnitude faster than on the EVM. Distributed compute on its own doesn't confer any points as a "store of value." The nature of the compute needs to be something that can be done strictly in the blockchain domain, and as we've seen, the result is a system that's easy to manipulate by spinning up new shitcoins, insider trading, wash trades, and worse - not to mention the loss of privacy given how effective on-chain analysis is. The features that make up a cryptocurrency make it decidedly unsuitable as the basis of a modern financial system. Some of the underlying principles are "neat," but that's as far as the tech goes for me.

[gus_massa]

I think it's worse. If you ask AWS for 1000 distributed machines, you can run 1000 different calculations and then somehow aggregate the results.

In the Ethereum network, if you ask for a calculation then all the miners must do it, and then all the validators must do it. So you acually have a repeated calculation, not a distributed calculation.

The only distributed part was calculating the next block with POW, but IIRC they switched to POS.

[wanderingbort]

I agree it’s worse on any measure that can reduced to “distributed computation”.

But that’s not what it’s doing. Not really. The only distributed computation is the one establishing trust that there has been no Byzantine failure of consensus.

Everything else is non-distributed non-parallel computation within a framework that statistically guarantees faithful and accurate processing.

It’s a competitor to trusted computing. One specifically designed against the single entity root of trust that underpins intels offering eg.

[edit: fixed some typos]

[ilyt]

That kinda doesn't matter when the amount of compute is trivially reproduced by '90's CPUs

You can just have a bunch of notes and vote out bad results. You can say "how do you know that 20 different machines you run that do not have hardware backdoor but how do you know software running ethereum is not backdoored ? There is no protection for the compute engine being backdoored, just the fact you'd need to get that backdoored software to most nodes.

[wanderingbort]

In order to convince one honest node you’d have to backdoor every node they talk to.

Multiple independent implementations co-exist on the network including many private implementations.

Perhaps the best defense is the vigilance of the community as they regularly build consensus around the head blocks hash which is hard to forge if the ledger contains any material impacts of an exploit.

Trivial detection of malicious acts is the feature.

It’s true it’s not perfect but your analysis trivializes the effort to compromise even one honest node.

[ilyt]

Or you can get your raspberry pi, some desktop from the 90s and random used laptop and get 1000x the "secure computing power" that couldn't be reasonably hardware-backdoored too.

Crypto for compute is entirely useless use case.

[barbariangrunge]

This thread is the only rational, technically informed discussion of crypto I’ve ever read. Both sides being polite to each other while making technical or philosophical arguments

[latchkey]

1. ETH switched off PoW to PoS, last year.

2. When ETH was PoW, miners were not doing any calculations. It was simply building a hash for a block template that a pool provided.

[alexambarch]

> You can already rent compute globally and perform computation in an environment orders of magnitude faster than on the EVM

True, but Ethereum also has the benefit of having a globally agreed upon state that is transparent and traceable when a change occurs. Not saying this setup couldn't be done any other way, but the fact that all you need to enable yourself to make those changes is a digital token is what I find interesting about it. The use of this token combined with the fact that people arguably care about the state of this system is what I feel gives it intrinsic value.

[threeseed]

The problem is the shared state is so small and expensive that the majority of the state gets managed elsewhere and is merely linked to the chain.

And so it's debatable if you've really achieved all that much.

[yunohn]

What does this state hold exactly?

[alexambarch]

What entity an ENS record is registered to, who owns what JPEG, who sent who some Ether...

Ethereum is one big state machine so that state holds everything.

[Yizahi]

For ENS - yes, for JPEG or whatever else - definitively no. There is literally nothing stored inside the token to describe who owns a digital or physical artifact outside of blockchain and it is technically impossible to store such info in the blockchain, while keeping even a current sham of decentralization, transaction throughput and gas price.

[nradov]

Ethereum (or any other blockchain) is unable to establish legal ownership of a JPEG.

[wanderingbort]

The interface to the globally available computation is a bit different and I think novel.

In order for my application and yours to interact we had to both chose the same ecosystem and invest in it to some minimal degree.

It’s easy to frame that negatively.

But there are some positive aspects to that mutual agreement to invest. Similar to how you may deploy an open standard of authentication inside your infrastructure, the expectation is that the community will improve tools and share innovations over time even if you aren’t directly interacting with any of them. It’s not required. It’s just an expectation.

Ethereum is like that. There is a community of developers extending the ecosystem and it’s exclusively people who invest in and use the platform.

We can mince words over the value prop of cryptocurrency and a decentralized ledger but, I don’t get the same belonging to a community on AWS or GCP.

FWIW most of the people I work with in the community dislike the same scammers that you do.

[mattdesl]

Placing control of the world’s compute under a small handful of powerful megacorps is not exactly ideal. Imagining a future where we can use and participate in (as stakers/validators) decentralized compute, storage, transactions, contracts, etc is pretty interesting.

[anonymouskimmer]

This blog post has a bunch of citations "showing that some assertion about the cryptocurrency ecosystem that crypto-bros make can't be true". And unfortunately decentralizing away from a small handful of powerful entities is one of these things. Below is one example.

https://blog.dshr.org/2022/09/impossibilities.html#more

"3) Impossibility of Full Decentralization in Permissionless Blockchains by Yujin Kwon et al (1st September 2019) provides a different formalization of the idea that economies of scale drive centralization by introducing the concept of the "Sybil cost":

    the blockchain system should be able to assign a positive Sybil cost, where the Sybil cost is defined as the difference between the cost for one participant running multiple nodes and the total cost for multiple participants each running one node.
    ...
    Considering the current gap between the rich and poor, this result implies that it is almost impossible for a system without Sybil costs to achieve good decentralization. In addition, because it is yet unknown how to assign a Sybil cost without relying on a TTP [Trusted Third Party] in blockchains, it also represents that currently, a contradiction between achieving good decentralization in the consensus protocol and not relying on a TTP exists."

[pavlov]

In the Ethereum PoS model, those with money perpetually own the network by definition.

At least with AWS, Amazon doesn’t also control the US dollar and the euro. If their service starts to suck, I can take my business elsewhere. But if Ethereum were to actually become a global standard, there would be no escape from the Ether plutocrats.

Any participation you and I can have in a system like Ethereum is table scraps from the overlords.

[anonymouskimmer]

> If their service starts to suck, I can take my business elsewhere.

This is an extraordinarily good point.

Centralization of Etherium means a centralization of both the VM and the currency. It would be as if the Fed and US Mint also owned AWS.

If Amazon sucks, as you say, you can take your dollars elsewhere. If Ethereum sucks, your Ether is going to be worthless, even if you could bridge it to another chain.

[pavlov]

I’ve sometimes thought that Ethereum represents the kind of computing network that 1950s Soviets would have appreciated. Massively inefficient, every operation needlessly repeated across every node only for the pretense of equality and democracy, and ideologically obsessed with a rhetorical illusion of immutability and permanence even though data can be wiped by simple agreement of the Politburo that controls the network.

“Now that the next five-year plan will be implemented as a smart contract on the Sovgosethereum, we’ll be mathematically guaranteed to meet our economic goals.” — Makes as much sense as the guaranteed crypto lending yields of 2021.

[mattdesl]

What you are describing does exist in some blockchains, like Tezos. But Ethereum's governance is social: not driven by token stake or "coin voting." In Ethereum PoS, users do not cede control of the protocol to validators - the users who run nodes and decide what software to run effectively own and control the protocol.

If the protocol starts to suck, users can change or fork the protocol (this has already happened multiple times, PoS being the latest example).

[threeseed]

Who exactly do you think underpins most of Ethereum. It's mega corporations. [1]

And decentralisation in crypto is in reality a myth when you have so many centralised players e.g. Binance, Metamask, OpenSea.

[1] https://crypto.news/3-cloud-providers-responsible-for-over-t...

[mattdesl]

The article cites a tweet, which cites Ethernodes. According to that site, 63% of hosted nodes are run by Amazon, or about 39% of all nodes. The next largest host is Hetzner at 7% of hosted (far lower than it was a few months ago, because they announced they are no longer supporting Ethereum) which accounts for about 4% of all nodes.

Not that surprising, since Amazon runs a huge percentage of all hosted compute in the world. Luckily, Amazon owning a large share of nodes doesn’t mean they own or can control Ethereum. But they certainly do pose a risk to the overall network security and health, and validation-at-home can and should continue to be made easier.

https://ethernodes.org/networkType/Hosting

[ilyt]

The ethereum DAO fork showed that the moment the "big ones" are hit there is not much stopping them from violating the "immutability" promise for their own gains.

Instead of few big banks there are few big mining pools deciding the fate, distinction almost without difference. And in both cases you need money to get in to the inner circle

[vba616]

Anarchy always falls apart. I don't know why anyone would think that's a historical accident rather than a logical inevitability.

[mattdesl]

There are no mining pools in Ethereum today. The chain is "immutable" only up to the point that the social layer wants it to be; majority users can change the protocol as they see fit (see EIP 1559, PoS, and other forks).

[nivenkos]

Most cryptocurrency mining still ends up being done by a few huge companies though.

[mattdesl]

Users cede very little control to PoS validators in Ethereum.

https://news.ycombinator.com/item?id=34685482

[yunohn]

There are innumerable companies you can rent cloud resources from - bare metal to “serverless”. Ethereum, on the other hand, is severely limited in what you can achieve for any amount of rent.