Wgaspar — you’re pointing at the right thing, and the flip-flop is the diagnostic.
Pollution from a diesel generator does not care whether the generator is powering a steel mill, a hospital, or a GPU farm. PFAS in groundwater behaves the same whether the cooling tower belongs to a chemical plant or a data center. If a concern is real, it does not change when the user changes. That is what “pollution is pollution” means, and you are right to notice that the position has migrated.
I think the migration tells us something about what was actually under the older objections. There are two distinct things that get called “environmentalism,” and the AI buildout has separated them in a way that the fossil-fuel debate did not.
The first is environmental stewardship — the recognition that we hold the natural world in trust and have an obligation to use it well. That position is older than the EPA, older than the modern environmental movement, and it is rooted in the same tradition that gave us “subdue the earth and have dominion over it” and the obligations that come attached to dominion. A steward who poisons the well he was given to keep is a bad steward. That ethic does not flip when the consumer changes. It applies the same to a coal plant, a refinery, an oncology infusion suite, and a GPU farm.
The second is environmental ideology — environmentalism as a political identity, where the position on a given pollutant tracks the political valence of the industry producing it. Coal: bad. Nuclear: bad. Then the AI capex flood arrives, the political valence of the industries shifts, and suddenly nuclear is acceptable, diesel backup is acceptable, and the locality’s groundwater stops being a topic. The position on the pollutant moved because the political identity required it to move. The pollutant didn’t.
That is exactly the test you’re applying without naming it: if the same person who objected to a coal plant on environmental grounds in 2014 is silent or supportive about a data center burning the equivalent fossil energy in 2026, the original objection was about identity, not about pollution. If they are still objecting in 2026 with the same vigor and on the same scientific grounds, the original objection was about stewardship.
My own report engaged Helen’s list with that distinction in mind. Two of her six concerns — diesel emissions during sustained generator runs, and evaporative water consumption in already-stressed basins — are real and apply regardless of the industry. The other four (low-frequency noise as a health hazard, sleep disruption from the noise, “stress and anxiety” framed as an environmental exposure, PFAS exposure to neighbors) do not survive scrutiny at the dose and distance involved. A stewardship objection should be able to make those distinctions. An identity objection generally cannot, because it was never really about the specific pollutant.
Thank you for the post — you put a finger on the right thing.