Every time a large language model answers a question, something burns. Not metaphorically. Data centers running the accelerator hardware that powers AI inference draw real electricity, generate real heat, and produce real carbon emissions. That’s been true for years. What’s changing is the scale.
SCIENCE GAZETTE
In 1980, wind power cost more than 55 cents per kilowatt-hour. Today that figure sits below 3 cents. Capacity factors, the measure of how much of a turbine’s theoretical output it actually delivers, have climbed from 22 percent for pre-1998 installations to nearly 35 percent now. Those numbers didn’t arrive on their own.
A refuse truck built from recycled fishing nets and wind turbine blades sounds like a design exercise that will never leave the exhibition hall. The ReEconic, unveiled by Mercedes-Benz Trucks at IFAT in Munich, is meant to prove otherwise.
Nuclear fusion runs on a paradox. To release energy, you need plasma hotter than the sun’s core. To contain plasma that hot, you need a magnetic field strong enough and stable enough to hold it indefinitely. The tokamak, a donut-shaped reactor chamber, is the best design physicists have come up with for doing that. It still isn’t good enough.
Fifty-four countries walked into Santa Marta this week with a shared problem and no agreed map. By the end of the conference’s first day, they had at least the beginning of one.
The venue was a theatre. The audience, government ministers. The message, delivered by scientists who’d flown to Colombia’s Caribbean coast, was that the window for an orderly transition away from fossil fuels is narrowing faster than policy is moving.
Somewhere in the architecture of EU policy, two offices are working against each other. One is writing energy strategy. The…
There’s a particular kind of concept car that exists purely to generate a headline, and then there’s one that tells you something real about where a company is heading. The Kia Vision Meta Turismo, making its global debut at Milan Design Week after a quieter reveal in Korea last year, feels closer to the second category.
Three years of failed experiments. Then, in late November, the results changed. Babak Bakhit, a researcher at the University of Cambridge’s Department of Materials Science and Metallurgy, had spent the better part of three years trying to build a memristor that actually worked at scale. Most attempts failed. The breakthrough, when it came, traced back to a single procedural change: adding oxygen only after the first layer had already formed. Small adjustment, different outcome entirely.
Europe didn’t discover its energy vulnerability in 2022. It had been warned for years. What changed was that the warning arrived in the form of a bill, and then another, and then a cold winter with gas reserves that might not last.
