Hacker News Reader

Comment by Minor49er | Hacker News Reader

It seems unusual to have a post that makes an environmentalist statement about computer energy consumption, but then end it by requesting donations via cryptocurrencies which are mined through their enormous consumption of electricity.

View on HN

It’s only unusual if you read things online without a healthy sense a humor.

Ethereum is moving to proof of stake, will hopefully solve that problem.

Ethereum has been moving to proof of stake since forever...

Ethereum was released in 2015. The first PoS cryptocurrency (Peercoin) was released in 2012.

Yes, but not yet and we're not 100% sure it will actually happen or will retain a lot of users. So if you care about computing efficiency, taking crypto (even ETH) now is a weird choice.

1. It’s already live on the PoS beacon chain, and has been since December 2020.

2. The specs for the transition are ready and it’ll happen next year. Target is before June. First long-lived Merge testnet goes live next month.

Yes, there's a live test. So as mentioned before: we don't know if it will happen for real transactions, if a meaningful number of people will accept it, and when will it happen. (It's likely, but still months away)

It's not a test chain, it's _the_ PoS chain that will be transitioned to. It's running on real Ether, over $30b worth of it. You're incredibly out of touch.

Sure, that chain is the target for the merge and it's running with real Ether. But it's still a test. Right now if PoS nodes say a transaction exists but PoW say it doesn't... then for all practical purposes it doesn't exist.

Someone recently told me that proof of stake is less energy intensive than proof of work, but it’s still pretty energy intensive.

Even if you assume 100 watts of power per validator, it’s a >99% reduction in energy consumption. In reality, the average power consumption per validator is probably closer to 1–20 watts due to the fact that additional validator don’t stress an existing system effectively at all.

Reading up on it some more, it does sound significantly less resource intensive, but it also sounds like migration from PoW to PoS is rather difficult and has some technical disadvantages. Is Bitcoin ever likely to migrate? If not, then it seems kind of moot, but perhaps new PoS players will come along and take over the market?

Just think of what it'd be like if Bitcoin was developed in the AI field and mining was only possible to implement in Python.

Would it be any different? The amount of mining done is related directly to the value of mining and inversely to the cost of hardware; the "efficiency" of mining a block doesn't matter directly. If you made it 50 times hard to mine a block (i.e. you reduced the hash rate by 50x), then you would most likely make each Bitcoin 50x more valuable because of the resulting scarcity. So if you could only implement mining in Python, exactly the same number of CPU / GPU cycles would be spent on mining as before.

Reply on news.ycombinator.com