> I would think my three bullet points should imply that I already know that energy is not a fungible resource. But there are still better uses than Bitcoin.
Okay, my apologies, I did respond a bit too glibly there.
> There is not a problem with shipping hydrogen from ghost cities in China.
There is the same problem with this that exists with many energy storage solutions: often the storage creates more carbon usage and other environmental effects than simply burning fossil fuels closer to the location where the energy is needed.
> Neither is there a problem with running datacenters that we need for other purposes than Bitcoin.
There are a bunch of reasons all the FAANG data centers are near population centers, but the two ones that are most obvious to me are:
1. Internet infrastructure is near population centers, because the people in population centers are the ones the internet it intended to connect to. If you want to hook into the backbone of the internet, you have to be located near the backbone of the internet. This is necessary for latency.
2. Data centers require technical maintenance, and technical workers capable of doing that maintenance tend to live in population centers.
> We are currently failing at combating climate change; but to have a hope of avoiding the worst case scenarios, fossils need to be replaced with renewables at an unprecedented scale. Not only current electricity use, but also energy used in transportation. It's a huge undertaking, where you need to ramp as quickly as you can in every way you can.
You're preaching to the choir here.
> The train is starting to finally move now, and in a few years, investment into renewables will not be the limiting factor. The limiting factor is how quickly it is possible to construct. Since the limiting factor is not investment; every kwh spent on Bitcoin is a kwh lost to combating climate change.
This is where we disagree, I think. There's a lot of limiting factors. Even if there's massive investment into renewables, there can always be more money and more efficient use of the money you have. One of the biggest inefficiencies of renewables is that they tend to be time-bound: as you probably know, wind produces the most at night, and solar produces exclusively during the day, and neither production cycle perfectly matches with the energy usage cycle.
(As you probably know), two popular areas of exploration here are storage and changing usage:
1. Every major storage solution either doesn't work yet, or is horribly inefficient in some way. Batteries are inefficient in output/input, and have downsides in terms of metal pollutants and disposal of retired batteries even when producing the batteries isn't the limiting factor. Gravitational storage requires huge amounts of space. Thermal storage simply doesn't work at scale yet.
2. Changing usage, i.e. running networked hot water heaters when other forms of usage are low, is a promising area of development, but so far the improvements in efficiency are very modest.
A third solution which I think you're maybe underestimating, is to simply produce more energy, and use the excess production at low usage times to mine bitcoin. The profits from this can be used for a bunch of purposes, for example to reduce the cost of renewable energy to consumers (thereby increasing adoption) or to build more renewable energy.
I'll add that while nuclear energy is not technically a renewable resource, it can be part of a strategy for reducing greenhouse gas emissions, which is arguably the most important aspect of renewables. Nuclear power solves a bunch of problems that renewables don't solve, but it also has the problem of mismatch with the consumption cycle: nuclear power production isn't cyclical like solar/wind, it's not adaptable: it's extremely expensive to spin up and down generation: there have been a bunch of times in Germany where it was cheaper for nuclear power plants to pay consumers to take their excess power than to spin down generation. Bitcoin mining has the potential to power consumption for nuclear power plants as well, which could greatly decrease the cost of running nuclear power.