Hacker News Reader

>But alot of the new non-fossil energy production is now added to the existing energy mix, not replacing fossil fuels.

This is a really good point that worries me quite a bit as well. Energy is essentially fungible so any energy used by wasteful and pointless endeavors such as cryptocurrency mining or model training is energy that's not going toward reducing our use of fossil fuels. Worse it's energy that still has externalities associated with production and storage such as mining for solar/batteries destroying natural areas and hydroelectric impacting river ecosystems.

Fossil fuels should be phased out ASAP and the way to do that is to stop increasing energy usage and to ensure renewable energy is used only for activities that are actually necessary such as food production and heating/cooling.

As someone working in the "cloud computing" industry I see companies spending huge amounts of energy to index non-production logs they'll never look at. I see developers wasting huge amounts of energy to create useless models for generating content nobody will ever look at. I see companies leaving huge infrastructures running 24/7 just in case someone might want to get an Ad served at 3am. And I see the enormous grey data center housing all that junk that now sits nearby where there used to be a pristine forest full of wildlife.

>Close by where I live (45 min drive), Google is building a huge server park that will require approximately 5% of Norways total electricity production.

It's amazing how much these things consume and at the end of the day they aren't doing anything of value. Communities really need to come together to prevent new data centers and remove existing ones.

>But alot of the new non-fossil energy production is now added to the existing energy mix, not replacing fossil fuels.

Do you know how replacement and phase out works in the real world? Hint, you don't just immediately install 100% new capacity and then rip out the still working old capacity.

Hint, electric companies boast about building new green infrastructure when its sole purpose is to satisfy growing demand rather than reducing existing reliance on classical electric sources. Which is misleading to the average person. I think you and the GP would probably agree on this point.

In the US the share has grown from 12% in 2012 to 23% in 2022. You might still have issues with the speed of transformation but your facts are not right. https://www.weforum.org/agenda/2023/03/us-electricity-energy...

Whose/what facts? I’m not talking about the US and neither is the OP or GP.

I know too little, that much is clear to me after I have started to look into this.

The situation may be somewhat unusual here in Norway, since basically all electricity is generated by hydro dams (~90%). We have made large strides the last two decades regarding energy savings (LED lighting, heat pumps and better insulation), and this has largely made it possible for us to transition rapidly from fossil cars to EVs while keeping our use of electricity quite stable.

What we are seeing now is a rapid expansion of mega datacenters that lays claim to extreme amount of electricity, more than these regions can steadily supply. This means that we need even more electricity on top of the hydro we have. This again means rapid expansion of solar parks and the destruction of vulnerable nature. Just the one Google datacenter now being built in my region will lay claim to ~5% of Norways energy use!

This worries me, and makes me doubt the "green" label on these projects.

> But alot of the new non-fossil energy production is now added to the existing energy mix, not replacing fossil fuels

I agree completely. Global CO2 output is still increasing, afaik, despite all the work on alternatives so far.

We need much more emphasis on using less energy.

I'm a fan of efficiency, but don't think "using less energy" is a particularly useful metric when that energy can come from sources with radically different costs and downsides.

Happily, the most effective way to use less energy is to electrify with renewables, which is a major part of the green transition

> I'm a fan of efficiency, but don't think "using less energy" is a particularly useful metric when that energy can come from sources with radically different costs and downsides.

That are often somewhat fungible. Using less is always good.

2024 is predicted to be the first year with a decrease in CO2 emissions.

Good luck fighting the AI compute gold rush. It's going to be worse than bitcoin mining, as these compute cycles actually do something.

> We need much more emphasis on using less energy.

This might be self-sabotaging in that it helps climate change deniers get elected. Politicians have finite political capital and degrowth is unpopular. Better use political capital for sustainable growth and accelerating renewables. Win-win narratives are the only thing that work politically.

Plus I don't think degrowth is even necessary. There isn't a zero-sum tradeoff between growth and emissions.

https://twitter.com/MaxCRoser/status/1771162980954079621

"Windmills" and electric cars do not, in fact, rely heavily on rare earth materials. Enormously more are used, e.g., in quadcopters. Even in places where powerful rare-earth permanent magnets are now important, they will soon be largely displaced by nitrogen-iron magnets, which are both radically cheaper and more powerful.

Google data centers are not examples of a "green energy transition".

How so? If it wasn't for the green energy transition, someone would buold a gas power plant there.

Hint: A Google data center does not generate power.

So what? Still, if it wasn't for the green transition, that power would be used, and a coal or gas power plant would be built to provide it.

There are so many years since commercial iron nitride magnets have been promised, but none have appeared, that I doubt that they would appear any time soon.

Like for many other inventions, it is likely that commercial iron nitride magnets would have appeared only if the original patent holders, who do not seem able to solve whatever technological problems exist that prevent their use, would have published which are the problems that block them and would have been willing to license the patents in advantageous terms to those able to solve such problems.

From the little published information it is not clear whether the iron nitride material is too difficult to produce or whether the crystalline structure is unstable in time, leading to a short lifetime.

> I am genuinely worried about the squeeze we are in now between conserving and protecting nature and transitioning away from fossil fules.

You shouldn't be, as pretty much the biggest thing you can do to conserve and protect nature is to transition away from fossil fuels.

Which you maybe already know since your example of a new data center has nothing to do with the transition from fossil fuels, and would only be much, much worse if their plan to power it was to burn fossil fuels.