It's hardly a rounding error, but I'm sympathetic to the fact that there are a lot more important concerns with respect to climate change than PoW, however, there is still a valid ecological critique of PoW due to it's inherent wastefulness relative to every other technology.
The amount of waste necessary to support the network must always grow since any new efficiencies are immediately obviated by the incentive to bring on more miners, the total utility provided by the network (i.e. the rate of transactions it securely processes) has no relationship to the amount of energy that the network burns. With every other technology, new effeincies make the technology able to do more useful work while burning less energy, in this way anything based on PoW is fundamentally flawed. You could hook up a fusion reactor of the future to the bitcoin network and it would not provide any more utility, yet the network would eventually consume all the energy produced by the reactor given enough time to increase mining capacity.
I know the typical response to this is PoS, and I think a switch to PoS would be great since its impact on the environment is within the realm of normal software. Whether or not PoS can actually work for a large network is a different discussion.
There’s a full featured testnet for the merge (Ethereum PoS) you can run right now. Yes, the developers were too optimistic with timeframes before, but it’s close and I’d expect it by EOY.
I'm following the progress. I don't like estimating release dates for other peoples' work, but I wouldn't be surprised if your EOY estimate is correct.
The biggest concern I have though, is that you can run as many testsnets as you want, but that doesn't mean the rollout is going to survive contact with the enemy. I'm very pessimistic that PoS can replace PoW in real-world usage. Once everyone is on the PoS ETH, I suspect that problems will eventually manifest, and either ETH will be forced to roll back to PoW, or there will be splits, and one of the PoW chains that splits off will supplant ETH in popularity.
I don't think this will happen immediately, but I think it's very likely to happen within a few years of ETH switching over to PoS.
The Ethereum PoS Beacon Chain mainnet (it’s not a testnet) has been running since December 2020: currently over 300k validator are active, with nearly 10 million “real money” ETH staked.
Yes, and it's definitely impressive, and I know this might seem like moving the goalposts, but the market cap of Beacon is very small compared to ETH Mainnet. More importantly, it's not _the_ ETH blockchain. Beacon may be working so well because PoS is optional. The people participating in Beacon have bought into PoS on a conceptual level and are working to make it work. When you're incentivized to, you can ignore the pretty fundamental design problems of PoS.
Once ETH tries to get everyone into PoS, that's where I think problems start.
If ETH weren't so established right now through the NFT marketplace, I might suspect you'd see a jump to other blockchains, like Bitcoin, that are still on PoW. Since it is established, I worry that people will try to make it work for a few years, but it ultimately won't work, and ETH Mainnet will either be forced to revert to PoW or lose popularity to another fork.
There's roughly USD $26 billion equivalent staked on Beacon, which puts it squarely in top-10 territory of crypto market caps, so I'm not sure why you consider it "very small", even compared to the ETH1 mainnet, but I guess "very/small" might be subjective.
While NFTs are a non-negligible component of trade volume on the Ethereum blockchain, numbers I've seen recently put monthly volumes of e.g. OpenSea in the single-digit USD billions equivalent. The daily volume of ETH is currently about USD $14 billion equivalent. I don't think the NFT marketplace is what's greasing the wheels.
I didn't mean to imply that NFTs were the biggest economic driver of ETH, but rather that they occupy the zeitgeist. There are systems that are successful because they are successful. If you look at Bitcoin, ETH, and a couple other altcoins, they are much more popular than all the other blockchains out there, and are seen as much more important than other blockchains. They have a perceived legitimacy and large numbers of people participating that other cryptocurrencies do not. Due to the nature of human attention and competition, there's room for a handful of cryptocurrencies in this space, and the rest, including Beacon, will never garner the same kind of attention or participation unless they replace one of the big ones.
When you have way more forces in the market, many who aren't motivated by making it work, I think that's where PoS has some real weaknesses, and I don't think we'll see practical attacks against it until one of the important ones goes PoS.
The amount of waste necessary to support the network must always grow since any new efficiencies are immediately obviated by the incentive to bring on more miners, the total utility provided by the network (i.e. the rate of transactions it securely processes) has no relationship to the amount of energy that the network burns. With every other technology, new effeincies make the technology able to do more useful work while burning less energy, in this way anything based on PoW is fundamentally flawed. You could hook up a fusion reactor of the future to the bitcoin network and it would not provide any more utility, yet the network would eventually consume all the energy produced by the reactor given enough time to increase mining capacity.
I know the typical response to this is PoS, and I think a switch to PoS would be great since its impact on the environment is within the realm of normal software. Whether or not PoS can actually work for a large network is a different discussion.