The whole thing reminds me of one Arthur C. Clarke short story, probably from the 1950s, where a group of Tibetan monks has purchased a computer to generate "all the one billion possible names of God". Before the computer, the monks had been doing the work by hand for centuries.

An American computer engineer is sent to install the machine. He chuckles at the monks' superstition: they claim that generating the billion gibberish strings is a task from God and the fundamental purpose of the universe.

They made a mistake by not increasing the search space (or difficulty in Bitcoin terms) in order to make that computer faith-compatible.

Then the Universe is destroyed when the billionth name is generated.

I'm not really in this world, but you'd save more power by recycling aluminium cans (%5+ of US power consumption) than by killing off cryptocurrency mining.

Personally I bring everything I can to recycling points, sometimes even if it's not my own garbage. But I'm aware people don't care.

On the other side the crypto mining is not only the power consumption: it's all the hardware and it's production as well. The ASIC antminers can't be used for anything else, hard to recycle as well and they constantly need to get bigger and faster. For Ethereum, people are buying shitloads of high end GPUs.

I also don't like the idea of storing insane amount of data, hoping to find correlations to make more money, but this is also nothing I can do anything about.

You understand that this article is about a cryptocurrency that uses a service (long term file storage and access) and the proof of work function instead of pointless computation.

Yes, I do, and this sounds like a good idea - emphasis on the sounds, because it's not new, see primecoin.

Filecoin at least isn't based in competition who can do "useless" computation the fastest, but computation only happens to verify data is actually being stored by those claiming they do so.