[eru]

I agree with most of what you say. However:

> [...] especially when carbon costs are priced in.

You misunderstand the economics of proof-of-work. Perhaps we should call it 'proof-of-waste'. Basically, in something like bitcoin proof-of-work functions a bit like an auction: the total amount of resources wasted on mining tends to equal the total amount of mining rewards.

For simplicity, assume mining rewards are fixed.

If hashing becomes cheaper, (eg first because of GPU miners, later because of ASICs), the hash rate will go up, but the total amount of resources wasted on mining will stay the same.

If hashing becomes more expensive (eg because of a universal carbon tax), the hash rate will go down, but the total amount miners spend on mining stays the same.

Arguably, that's exactly how a carbon tax is supposed to work in the best case.

Of course, there will be plenty of incentives to avoid the carbon tax. Both by mining with clean power (good!) and by mining in places that don't implement the carbon tax (less good!).

[spinny]

> proof-of-waste

half agree with that. PoW is the most efficient method of converting raw energy into a digitally secure token. It would be much better if the work done was actually useful beyond that.

Some attempts have been done in the past, but with low success (Riecoin IIRC uses cpus to find sequences of primes, GridCoin used BOINC for the PoW, proof-of-boinc in that case).

There is still work to be done on EVM chains. Contract calls need to be processed and verified by every (full) node

[eru]

> half agree with that. PoW is the most efficient method of converting raw energy into a digitally secure token. It would be much better if the work done was actually useful beyond that.

I'd say PoW is not the most efficient method. It's a terrible waste by design.

The kind of 'work' that works in PoW has some rather strict requirements, that's why it's almost impossible to shoehorn useful work into a PoW scheme.

Wikipedia tells me that eg GridCoin uses proof-of-stake to prevent double-spending attacks; its rewarding people for BOINC work is a totally separate gimmick.

[lubesGordi]

Proof of waste is an interesting way of putting it. I don't disagree. Maybe it would be interesting for power companies and govt to pair up and provide coins that are deflationary, and are only mined by putting power back into the grid. Make those coins useful to pay for car registration, taxes, or whatever. At least some deflationary value could be put to work there.

[eru]

> Maybe it would be interesting for power companies and govt to pair up and provide coins that [...] are only mined by putting power back into the grid.

That idea sounds good, but would only be a solution for a problem that doesn't exist.

Bitcoin's design solves two problems:

(1) only someone who owns a coin should be able to spend it

(2) they should only be able to spend it once

The first problem had been solved for ages: use digital signatures to authenticate messages of essentially the form "I, Alice, hereby send 3 bitcoins to Bob." Given some initial coin assignment to specific private/public key pairs, you can follow along these messages, keep a tally in a ledger, and know how much everyone has.

Solving the second problem is where bitcoin innovates. Without a trusted third party coordinating things, Alice could just show a message to Bob and Charles each 'sending' them the same 3 bitcoins. Bitcoin uses proof-of-work to pick (at most) one of these messages to be part of the consensus reality.

The scheme you are suggesting requires some trust in the partnership of government and power companies. But as soon as you have very minimal amounts of trust, you can just solve the double-spending problem directly.

And that direct solution involves much less trust than what is required for accepting any statements about the power grid.

If you want, I can sketch out a minimum trust system for the direct solution I mentioned.