"...normalized them so people treat them as permanent unchangeable background parts of the world. If we want to think seriously about climate, we need to be willing to slice through those illusions and think of the whole world as more malleable"
I think this is the real point and data centers are the flavor of the month. For whatever reason people just seem bad at context and calibrating. Every endeavor we take on has pro and cons and those should be viewed against realistic alternatives and the cost of doing nothing.
Very true. Just a thought but in my experience one way the microwave/computer analogy breaks down is that a lot of people, especially over 40s, believe we should stop using computers altogether. They'd happily wind the clock back to 2003, after Google but before social media. Of course their revealed preference suggests something very different but the explicit belief that we could do without the extra compute enables some of the hostility towards data centres
Data centers are perceived as a particularly defiant and provocative symbol of "tech we don't like by people we don't like, in your face", just like the national microwave would be.
As you pointed out before, some people seem to think expending energy on non-physical things is wasteful by definition. Constructing massive AI-centric datacenters would be the epitome of that, _plus_ signaling that this undertaking isn't transient fad or subculture hype. Data centers are about openly planting roots and building infrastructure for the long haul. They represent an actual change made in the physical world, and I can see how that might feel infurating if you want AI to simply go away. There's no hope of soft-canceling concrete.
All good points, a few replies: First, I really don't think AI is only being used for unnecessary crap, and it takes up a very very small portion of the unnecessary ways we use energy. If you think there's more positive chance for AI to decrease emissions in other areas (in the same way the internet does) then it starts to look more reasonable. Second, I don't mean this to say "These emissions literally don't matter at all" only that we need to consider them in the big picture and prioritize the places where we can actually cut the most the fastest. I don't think data centers will be anywhere near the top of that list. There are a lot of other places we're adding a ton of unnecessary energy (increasing meat consumption etc.) that I think should get a lot more attention. Finally, I don't see this as a race to the bottom argument. I think instead I want us to ask "Regardless of which industries are new and old, we need to mainly look at what's actually going to cause the greatest reductions in emissions, and where the most promise is to do that." Just focusing on cutting industries because they're new doesn't seem like a good argument, especially if they have a lot of potential to be useful for the environment (as I think both the internet and AI do).
"...normalized them so people treat them as permanent unchangeable background parts of the world. If we want to think seriously about climate, we need to be willing to slice through those illusions and think of the whole world as more malleable"
I think this is the real point and data centers are the flavor of the month. For whatever reason people just seem bad at context and calibrating. Every endeavor we take on has pro and cons and those should be viewed against realistic alternatives and the cost of doing nothing.
IMO, it is about hating big companies. Make them the badguys. Us vs Evil Capitalism!
Etc.
Very true. Just a thought but in my experience one way the microwave/computer analogy breaks down is that a lot of people, especially over 40s, believe we should stop using computers altogether. They'd happily wind the clock back to 2003, after Google but before social media. Of course their revealed preference suggests something very different but the explicit belief that we could do without the extra compute enables some of the hostility towards data centres
Data centers are perceived as a particularly defiant and provocative symbol of "tech we don't like by people we don't like, in your face", just like the national microwave would be.
As you pointed out before, some people seem to think expending energy on non-physical things is wasteful by definition. Constructing massive AI-centric datacenters would be the epitome of that, _plus_ signaling that this undertaking isn't transient fad or subculture hype. Data centers are about openly planting roots and building infrastructure for the long haul. They represent an actual change made in the physical world, and I can see how that might feel infurating if you want AI to simply go away. There's no hope of soft-canceling concrete.
All good points, a few replies: First, I really don't think AI is only being used for unnecessary crap, and it takes up a very very small portion of the unnecessary ways we use energy. If you think there's more positive chance for AI to decrease emissions in other areas (in the same way the internet does) then it starts to look more reasonable. Second, I don't mean this to say "These emissions literally don't matter at all" only that we need to consider them in the big picture and prioritize the places where we can actually cut the most the fastest. I don't think data centers will be anywhere near the top of that list. There are a lot of other places we're adding a ton of unnecessary energy (increasing meat consumption etc.) that I think should get a lot more attention. Finally, I don't see this as a race to the bottom argument. I think instead I want us to ask "Regardless of which industries are new and old, we need to mainly look at what's actually going to cause the greatest reductions in emissions, and where the most promise is to do that." Just focusing on cutting industries because they're new doesn't seem like a good argument, especially if they have a lot of potential to be useful for the environment (as I think both the internet and AI do).