[22c]

So it seems Drew is biased here, he is biased because his service is being used by thieves to make money.

My tainted opinion: I think the criticisms of Proof-of-Work are totally justified here. There are perverse incentives for "miners" to execute work in the cheapest way they can, whether that is JS miners, bot farms, abuse of free services, hacked AWS accounts, abusing favorable tenancy conditions etc. and even if the miner isn't downright stealing, there's a good chance they are operating in an environment where the cheapest electricity isn't going to be exactly "green" either.

Practically all of these mining efforts are not done out of some kind of altruistic, utopian goal of shutting down central bankers, but are simply done to make money.

What I think is unfair of Drew is to categorize cryptocurrency as a whole as an abject disaster, when in his own article he writes "attempts at reform, like proof-of-stake, are viciously blocked". Well proof-of-stake still falls under the cryptocurrency banner, doesn't it? Perhaps it's better to say that cryptocurrency is facing an identity crisis.

My opinions on proof-of-storage: I always expected that the price of these coins would naturally level out because you'd have to remain competitive with the likes of Backblaze or S3. I don't know enough about the incentive structures of Proof-of-Storage coins, but if there are incentives to have empty spinning rust then these are the wrong incentives and I'd say those protocols are likely broken. Rewards should be heavily slashed if there's wasted storage (ie. empty storage), so as to disincentivize miners from over-provisioning hardware (hoarding disks).

Disclosure:

I'm currently holding roughly a pay check each of Harmony ONE and Polkadot and have a small, locked stake in ETH2. All of these projects are Proof-of-Stake.

I also own pocket change amounts of Stellar Lumens which are used between myself and a friend to keep track of who's paid for lunch. As I understand it, the Stellar consensus protocol is not computationally intensive when compared to Proof-of-Work.

[ddevault]

>So it seems Drew is biased here, he is biased because his service is being used by thieves to make money.

Thieves using my service to make money is what pushed me over the edge into writing this piece, but I have a well-documented history of criticising cryptocurrencies. Here are some examples on HN:

https://hn.algolia.com/?dateRange=all&page=0&prefix=true&que...

[22c]

You're asking people to document their own bias, so I think it is only fair to point out yours.

Do you believe that cryptocurrency as a whole is an abject disaster? Do you support efforts to reform consensus with alternatives to Proof-of-Work (eg. Proof-of-Stake, Delegated Byzantine Fault Tolerance, etc.) or do you think cryptocurrency is a lost cause?

[ddevault]

>You're asking people to document their own bias, so I think it is only fair to point out yours.

I think I have been pretty open with my own perspective in TFA.

>Do you believe that cryptocurrency as a whole is an abject disaster?

Yes. I just published an article titled "Cryptocurrency is an abject disaster".

>Do you support efforts to reform consensus with alternatives to Proof-of-Work [...] or do you think cryptocurrency is a lost cause

I addressed both of these in TFA, too.

[22c]

>I think I have been pretty open with my own perspective in TFA.

Perspective and bias are two different things.

>I addressed both of these in TFA, too.

No you didn't, or I wouldn't have asked. You mention proof-of-stake in passing. You also seem to continue to conflate proof-of-work with cryptocurrency in general.

You seem uninterested in discussing these topics further than what you've laid out in your article, which is totally fine. Thanks for taking the time to write and share.

[majewsky]

> I always expected that the price of these coins would naturally level out because you'd have to remain competitive with the likes of Backblaze or S3.

Backblaze and S3 cannot conjure storage space out of thin air. If proof-of-storage coins make hard drives more expensive, that also increases the capex for those services and thus their prices. In the end, they're just shitting on everyone's lawn, no matter if you're using your own purchased disks or storage-as-a-service.

Stake: I hold EUR and have never held crypto"currencies".

[22c]

>Backblaze and S3 cannot conjure storage space out of thin air.

The same goes for Proof-of-Storage miners, right? Whoever has the most buying power wins.

[danaris]

I've heard a lot about proof-of-stake as being a lesser drain on energy per unit of mining, but I'm still very unclear on—well, many things about PoS, actually, but one in particular stands out:

How would eliminating all Proof-of-Work cryptocurrencies and shifting entirely to Proof-of-Stake eliminate the incentive for miners to monopolize every last bit of computing resources they can?

My understanding of cryptocurrencies and blockchains is still fairly limited, but I'm struggling to see how any system that allows for conversion of computing resources into a saleable commodity can avoid that problem, whether or not it finds a way to avoid the "electricity-consumption-to-money" pipeline.

[22c]

>How would eliminating all Proof-of-Work cryptocurrencies and shifting entirely to Proof-of-Stake eliminate the incentive for miners to monopolize every last bit of computing resources they can?

I guess the point here is that you could effectively stake on a Raspberry Pi with a 256GB SD card for the next 5 years or so, and you don't need more Raspberry Pis or SD cards to stake higher amounts. The main limiting factor is typically you need to be a full-node, which means syncing the entire blockchain.

If that's not really green enough for you, you can also do delegated staking. One person could run a staking service on their Raspberry Pi for 1,000 people, etc.

[danaris]

...Can you expand on that a little further? I don't really understand what it means "to stake" in this context, or why more RPis (or other compute resources) wouldn't let you mine more/faster.

Or if you have a link to a good resource for explaining how proof-of-stake works in a way that would clarify this, that would also be fantastic.

[22c]

> more RPis (or other compute resources) wouldn't let you mine more/faster.

Essentially, because there is no computationally intensive work being done. Proof-of-Stake is kind of like saying:

"I have 100 of these coins staked in the network that I can't spend, because of my faith in the network, I'd like to validate this other transaction of 10 coins, the network should award me with 1 coin for validating this transaction because I'm being a good validator.".

Another validator has an incentive to prove that those validations are invalid, because if they do, they get the reward instead, and usually an additional fee.

"Not so fast, I have a copy of this ledger as well, and I can see that this transaction could not be possible. I also have 1000 coins staked, so the network should trust me even more than it trusts you! Not only should I get the reward instead, but we should all agree that you'll get no more rewards in future until you clean up your act."

The rest of the validators, along with their stake, then essentially vote on who wins the dispute. All of this is not computationally expensive. The person staking 1000 coins and the person staking 100 coins can do so both on the same hardware, but yet one of them has more "voting" power on the network, and can net more rewards by having more validation throughput.

That's all very over-simplified, but I think it covers the basics.

[danaris]

Thanks for replying, even so late!

If you don't mind, then, can you explain what does allow for staking more coins at a time? That is, if I wanted to increase my rate of mining by 100x in a PoS system, how would I do it?

[22c]

That's simple, you just stake more coins. Staking rewards go back into the stake, so it is compounding (ie. you stake 100, get a reward of 1, you are now staking 101). As long as you don't unstake, your stake will grow over time.

If you want to stake more coins than you have, then you need to acquire coins by other means (ie. buy them off someone who doesn't want to stake them).

Stakers also have an incentive to trade some of their rewards instead of re-staking them, because they want the network to get utilised rather than simply be 100% staked.

[davidgerard]

> How would eliminating all Proof-of-Work cryptocurrencies and shifting entirely to Proof-of-Stake eliminate the incentive for miners to monopolize every last bit of computing resources they can?

It absolutely wouldn't - as we see from Proof-of-Space trashing the hard disk market recently.

[22c]

Proof-of-Stake != Proof-of-Space

Also Proof-of-Space != Proof-of-Storage (seems a lot of people are conflating the two in this thread as well).