theSIGNAL SCIENCE
28 August 2026
"Technology is neither good nor bad; nor is it neutral."
Science

The Noisy, Thirsty Neighbours Nobody Voted For

AI's invisible engine is a very visible problem — and the communities living next to datacentres are starting to fight back.

Picture a quiet village on the edge of a British town. Then picture a vast, humming concrete box appearing beside it almost overnight. No one asked the neighbours. No one took a vote.

That is the reality that communities across the UK — and on every continent where AI infrastructure is expanding — are beginning to confront. Datacentres, the physical buildings that make artificial intelligence run, are arriving fast, and they are not quiet guests. The noise is one complaint. The sheer ugliness is another.

But the deeper grievance is environmental. Datacentres consume extraordinary amounts of electricity and, in many designs, enormous quantities of water for cooling. When governments promise rapid AI growth and clear the path for developers to build, local residents often find that their objections carry very little weight. Planning rules have, in several cases, been adjusted specifically to speed construction along.

Photo: Craftsman Concrete Floors / Unsplash
A large datacentre facility sits on the edge of a residential area, its cooling units running continuously.
Discuss
  • Why might governments prioritise datacentre construction over the objections of local communities?
  • Is it fair to say that AI's environmental costs are 'invisible' to most users — and does that matter?
  • Should planning laws treat datacentres differently from other large industrial buildings, and why?
Science

AI cracks a 100-million-option puzzle

Imagine testing 100 million combinations — one by one. Each attempt costs hundreds of dollars, takes days to analyse, and sometimes ends with a melted mess. That was the reality of 3D printing GRCop-42, a copper alloy NASA engineered for rocket engine combustion chambers. Researchers at Washington State University trained an AI to skip the guesswork. It identified workable printing settings that standard commercial machines — which currently fail 90% of the time with this alloy — could actually use. "We are essentially democratising the printing of this alloy," said computer scientist Jana Doppa. The same strategy could now be pointed at drug discovery.…
  • Should AI-driven research tools be openly shared across all countries?
The technology may be borderless, but its costs land somewhere very specific.
Aisha Down, Global Technology Reporter, The Guardian
Science

Fired for Teaching Climate Truth at West Point

For 17 years, Dr Adam Kalkstein taught army cadets at West Point about the human causes of climate change. Then, in November 2024, a general told him to stop — and he refused. Kalkstein, West Point's only climate scientist, called the order "immoral and potentially illegal." He compared it to teaching law while banning any mention of the constitution. The general stormed out, citing Donald Trump as commander in chief. Kalkstein was fired after the spring semester. His lawyer describes him as a conservative Republican who believes cadets simply cannot defend a nation whose climate threats they are forbidden to understand.…
  • Should military academies follow government policy or scientific consensus in classrooms?
AMERICAS · Science
Manhattan-sized ice island breaks off Greenland’s Petermann Glacier
WORLD · Science
Shingles vaccine linked to lower risk of heart attack and stroke, study finds
ASIA · Science
Dozens of US and Canadian citizens missing after flash flooding in Nepal
OCEANIA · Science
Bird flu is spreading. Your backyard feeder could make it worse
theSIGNAL IN THE LAB
1VOCABULARY
grievancepushback
democratisingdownwindconfront
objectionsworkable
2GRAMMAR FOCUS
Wish and if only — expressing regret
Use 'wish' or 'if only' + past simple to express regret about a present situation, or + past perfect to express regret about a past situation that cannot be changed.
had been consulted · could understand · had refused · were · had not been fired · knew · had listened
  1. Residents near the new datacentre wish they before construction began.
  2. If only the planning rules stricter, communities would have had more protection.
  3. Dr Kalkstein wishes he — but his principles made that impossible.
  4. Many cadets wish they the full truth about climate threats facing their country.
  5. If only the general to Dr Kalkstein's argument before storming out of the room.
  6. Researchers at Washington State University wish it easier to 3D print GRCop-42 without AI assistance.
3DEPENDENT PREPOSITIONS
Complete each phrase with the correct preposition. All from today's articles.
  1. Communities are beginning to confront the reality ___ AI infrastructure expanding into their areas.
  2. Residents found that their objections carried very little weight ___ the planning process.
  3. The technology may be borderless, but its costs land somewhere very specific — ___ communities downwind.
  4. Researchers trained an AI to skip the guesswork ___ identifying workable printing settings.
  5. Kalkstein compared the order ___ teaching law while banning any mention of the constitution.
4CRITICAL THINKING
Dr Kalkstein argued that soldiers cannot defend a nation whose climate threats they are forbidden to understand — yet the AI datacentres powering military systems are themselves significant contributors to carbon emissions. Does this contradiction reveal something deeper about how institutions handle inconvenient truths? Who, ultimately, is responsible for ensuring that those in power act on accurate information?
5CREATIVE · HEADLINES
Write a headline for the top story in each of the following styles. One line each, no explanation:
  • TABLOID NEWSPAPER
  • LUXURY MAGAZINE
  • ACTIVIST BLOG
6WRITING
Using evidence from all three articles, argue either that technological progress consistently overrides the rights of individuals, or that individuals still retain meaningful power to resist it — even in the age of AI.
7DEGREES OF EXTREMITY
Complete each ladder from mild to strong.
  • noisy
  • concern
  • uses
  • dislikes
  • unusual
  • unhappy
8SPEAKING
  1. Can resistance to datacentres unite unlikely political allies?
  2. What responsibility do AI companies have to communities near their infrastructure?
  3. Does refusing an immoral order make someone a hero or a liability?
  4. Could AI ever be used to solve problems it has helped create?