💾 Archived View for idiomdrottning.org › proof-of-work captured on 2022-03-01 at 15:07:16. Gemini links have been rewritten to link to archived content

-=-=-=-=-=-=-

One hand knows what the other hand does

The Swedish gov’t is gonna contribute 6000 million SEK to people’s electricity bills. Probably a needed thing, short term, but:

1. If even one fraction of the reason for them having to do that (the high electricity cost) is because of miners (driving the price up (even partially as one of many factors)), I’m salty about that.

2. If even one fraction of the outcome of them doing that (alleviating the high electricity cost) is gonna go to mining, I’m salty about that too.

3. The Swedish state, as pushed for by their state-owned electricity-selling company through their state-owned media outlets, are promoting PoW mining, PoW coins, and NFT. I’m super salty about that.

It’s not difficult to understand.

People pour in power (and [often shockingly short-lived] hardware) to get more chances at creating successful blocks.

PoW is maximally inefficient by design.

In bitcoin/src/pow.cpp:

if (UintToArith256(hash) > bnTarget)
    return false;

In other words, if the hash is over the difficulty limit, the checker indicates that it should get discarded. Meaning that the more juice people destroy, the more more juice the next batch has to destroy. (And the bnTarget can lower, as it did when China cracked down on mining.)

The amounts of blocks is constant (2016 per fortnight), it’s just a race to the bottom about who can throw the most power out the window to get them.

So mining is a bad idea where electricity is expensive, which is why so much mining is on cheap hard coal in Kazakhstan. Which is also a bad idea.

Electricity sellers are happy to sell to miners, which is part of the problem here in Sweden where, again, the state-owned electricity seller is saying messed up stuff like “mining can ‘buffer’ electricity usage on non-peak hours” (wack since, unlike a battery, as inefficient as batteries are, it‘s not possible to get electricity back from blocks).

Blocks don’t generate value. Their only value as an investment is to offload them to bagholders down the line. They do have a use value as a transactional tool—like checkbooks—but compared to other transactional tools, blocks are several orders of magnitude more inefficient.

“We are gonna save the forest by chopping down trees during peak growth hours, make paper, print money on that paper, and then buy new tree saplings for that money. This is buffering.” That doesn’t make any sense. The tree saplings aren’t created from the money. The money doesn’t have inherent value beyond its system-use as a transactional tool.

And, no matter how much power we throw away, we’re not getting more blocks total. In other words, we’re not even contributing more transactional use value since there would’ve been 2016 blocks anyway (just that someone else would’ve gotten the reward for it).

In the tree-chopping analogy, it’s as if when we print money we would also destroy an equal amount of bills somewhere else since it’s a constant-sum game. One block per ten minutes. This isn’t buffering. It’s just waste.

Transactional tools are necessary in market capitalism, but we should use efficient transactional tools. Now, are there benefits to verifiably scarce, anonymous, digital scrip? Sure, such as buying illegal drugs and paying ransom money. Hopefully humanity can figure out a non-PoW way to fill that niche instead. (We’ll see if PoS, FBA etc can pan out. They have some other risks.)

Because PoW is the worst invention of all time.

Update

Here is a fancy overview.

(Older stuff, from 2021-10-01, follows.)

Proof of work

Chris mentioned this six-month old thread on Hacker News.

I don’t sign off on his “rot in hell” sentiment in there, but that thread had plenty of messed up stuff.

For example, arcanon wrote:

If crypto increases energy demand and new energy capex that is most efficient is as green energy, then you are all wrong and crypto is good for the environment.

That’s not a great way to look at it. Remember that renewables doesn’t mean infinite, especially in terms of bandwidth. Särimner needs some time to regrow.

We want less fossil fuel burning. Greenhouse gases going from lithosphere to atmosphere is the root cause of catastrophic anthropogenic global warming.

Spending energy on proof of work is not decreasing any fossil burning. It’s just not.

As an analogy, let’s say I want my students to get safe jobs instead of risking their lives as guillotine testers and tightrope dancers.

I take a quick glance at the jobs section and see that hmm, safe jobs seem to be paying better over the long run, if not in the short run.

So I take my students’ paychecks and burn the money in a big old firepit in the yard.

They’re like “Miss, miss, why’d you burn all that cash!” I’m like “If I waste all your money, I’ll increase your demand for money, and the new money capital expenditure that is most efficient is gonna be safer jobs” they’re not gonna go “thank you miss! That makes sense! We’ll start working those safe desk jobs right away so you can pour more of that money in the firepit!“

Instead, they’re gonna go “OMG we’re starving and we need cash right now and all our money got burninated, we are gonna go have to go work as dynamite jugglers!”

They’re not gonna go “Thank you for saving money on paper by flushing the groceries directly into the toilet.”

arcanon went on:

Not to mention that it directly couples energy and monetary value.

In a pretty messed up way. Here’s another analogy, let’s say I invent a machine where you pour in clean good drinking water and the machine utterly destroys the water and then spits out twenty-dollar bills. That would directly couple drinking water and monetary value, in an horrific way. That is what proof-of-work is doing.

knuthsat in that same thread wrote:

There is nothing wrong with proof of work. It is useful. People also use Bitcoin, Ethereum and other cryptocurrencies for useful things. Energy demands are nothing compared to other useful human activities.

Also, energy consumption for Bitcoin rarely has other externalities.

Livestock based diet is another collective human activity that wastes enormous amounts of energy, creates disastrous amounts of pollution (water, air, ocean...).

Proof of work, compared to that is really insignificant.

I argue against animal agriculture all the time. Easily the easiest way for humans to significantly cut emissions is to ditch the animal ag. It’s not gonna be enough but it’s a really big low hanging fruit. Please, eat plants.

That doesn’t make it correct to say that proof of work’s energy demands are “nothing compared to other useful human activities”. Look up the numbers before guessing. They are huge. That goes for both the direct use of power, and the mining for metal for the single-purpose, disposable electronics used to mine.

Listen, I get it. What the thinker thinks, the prover will prove and it’s difficult to change your track on this stuff. I thought crypto was awesome too, first I heard of it, having long been frustrated with the centralization of stuff like Flattr or Paypal. Also, we need the crypto to buy drugs. And, these protocols are interesting to think about.

No matter how much you like it, that doesn’t mean you can just try to wish away the environmental impact, which is legitimately huge.

The Bitcoin White Paper

Three years ago, Scott McCloud listed seven drawbacks of traditional money:

1. Prone to inflation and mismanagement

2. Increasingly not private

3. Too easy to hack

4. Splintered into 180 slow-to-transfer local currencies

5. Hard to divide into small increments

6. Plagued by growing transaction fees

7. Largely inaccessible to about two billion people

Proof-of-work cryptocurrency only address points two and five out of those and are way worse on some of the others.

Cryptocurrency is very splintered, the transaction fees are pretty sick, and wow, talk about inaccessible.

Energy Use Compared to Banking

Nytpu writes in with an interesting comparison.

Yeah, it’s pretty bad. That’s not counting all the e-waste, either, which is an even bigger problem than the direct electricity use.

Energy use compared to Chrome tabs

There’s also the energy use compared to Chrome tabs argument.

I hope someone does do the math so we’ll know for sure. Those heavy JS pages needs criticism, for sure.

However, most people don’t participate in or benefit from the proof-of-work cryptocurrency pyramid scheme. Arguing for the widespread adoption of proof-of-work is like arguing for filling the streets with airplanes.

The Difficulty Variable

gerikson tells it straight.