Comment by stavros
4 hours ago
We're talking about global warming specifically here, though. Cars and planes should be a much bigger worry than AI power usage.
4 hours ago
We're talking about global warming specifically here, though. Cars and planes should be a much bigger worry than AI power usage.
Not when AI is directly resulting in increased greenhouse gas pollution. It's all of the above. Any source of greenhouse gas pollution is bad. Cars, planes, ships, AI data centers running on fossil fuel energy. It's all bad.
No. This is disingenuous. Something that consumes electricity doesn't care where the electricity comes from. Fix the power source, and you automatically fix every single consumer in existence at once.
I think your comment is the disingenuous one. We have no time left and "Fix the power source" is happening way too slowly in the real non-theoretical world. But what can happen in zero time is to not build another data center for something that nobody really needs.
narrowing the topic, that is exactly the quality that energy transition theorists are leaning on. The electrical grid is uniquely able to maintain a stable engineered and market place while inputs and loads change quite a bit.
There's an easy 19th century solution to cars and planes - public transport. It could reduce the usage significantly, save people lots of time, reduce pollution, make people healthier through making the environment more walkable, reduce crime. We don't do it not because the technology isn't there, but because it's more profitable for people to induce consumption by planning our cities and suburbs around cars.
There's lots of rotting low hanging fruits ignored for decades because politicians are paid by the ladder-sellers.
> save people lots of time
Public transit is rarely a time saver for people who give up their cars in favor of public transit.
> reduce crime
In what way? Car break-ins presumably go down when there when fewer cars, but does overall crime drop? Doubtful.