Comment by gruez
5 months ago
This video is unconvincing. The thesis of the video basically boils down to "AI companies are using electricity, driving up prices, and residential consumers are paying for it". However it neglects that most of the power usage is caused by the same residential users, who are getting something out of it. 80–90% of AI energy usage is estimated for inference (eg. generating responses on chatgpt), not training[1]. Moreover despite AI companies being unprofitable as a whole, they're making fat margins on inference[2]. Therefore it's reasonable to assume that most of AI electricity consumption that the video complains about is as a result of ordinary people using AI. Taking this into account, blaming AI companies for jacking up electricity prices in this context makes as much sense as blaming airlines for jacking up oil prices. They're only doing so because people are buying their services, and presumably deriving some sort of utility from it.
[1] https://www.technologyreview.com/2025/05/20/1116327/ai-energ...
[2] https://martinalderson.com/posts/are-openai-and-anthropic-re...
You are not taking into account the fact that data centers tend to be pretty concentrated geographically.
If I live in an area that is trying to attract data centers by giving them sweetheart deals, my electricity bill is going to go up by more than my share in inference usage since those data centers will be used by people all over the globe.
Many, people, probably most, are not using any inference in their lives - certainly not willingly or knowingly.
Worldwide, maybe. But in the "western world"? Couple weeks ago my mother asked me to compare the main LLM providers. And then for tips how to best use it. She is over 80.
1 reply →
Of course, for example me, the person you are replying to. I am an LLM vegan.
How is this different than any other form of industry?
If I live in an area that is trying to attract [steel mills] by giving them sweetheart deals, my electricity bill is going to go up by more than my share in [steel] usage since [that steel] will be used by people all over the globe.
I think if you found the average person on the street and asked them if their electric bill would go up when a steel mill opens nearby they'd only say yes because why else are you asking them.
Most other industries create (more) local jobs which would offset the localized cost.
This sounds like a generic NIMBY argument to me? eg. "why should we build a datacenter in my backyard, when it'll raise MY electricity prices?" Of course, if everyone has this attitude, no datecenters, power plants, homeless shelters will ever be built
I am very YIMBY, but it's totally reasonable that you don't want industrial work done in your backyard. This is exactly what zoning and city planning were designed for.
1 reply →
Then talk to your local politician
Very anecdotal, but a friend of mine works for a fuel cell company that basically converts natural gas and/or propane to grid electricity a huge part of their work right now is powering new datacenters for AI growth. I don't really know the technology, so I may have the details wrong, but the fact companies willing to take on cost and complexity does hint towards real limits on energy supply.
> Taking this into account, blaming AI companies for jacking up electricity prices in this context makes as much sense as blaming airlines for jacking up oil prices.
I'm not sure if "blame" and "jacking up" are the right words here. AI technology is demanding power quicker than it is being added to the grid and add so there is going to be tension. Now if tech companies start buying the grid operators and power, I'm more inclined to start blaming the tech companies themselves, but for now I still place broad energy failures squarely with the government.
I don't use AI.