The United Nations has put numbers to what many in the energy and technology industries had been calculating quietly. By 2030, AI data centres will consume 945 terawatt-hours of electricity annually, a figure that approaches the combined national consumption of Pakistan, Bangladesh, and Nigeria. Their water footprint will equal the basic annual drinking needs of 1.3 billion people.
Few events in astrophysics rival the violence of a core-collapse supernova. In a matter of seconds, a massive star exhausts its nuclear fuel, its core implodes under gravity, and an explosion powerful enough to outshine entire galaxies erupts into space. Yet despite decades of observation and theoretical work, the exact mechanism that determines whether such a star successfully explodes or collapses into silence remains unresolved.
For years, the universe appeared to whisper in a predictable rhythm. Across the frozen silence beneath Antarctica, the IceCube Neutrino Observatory recorded streams of ghostlike particles arriving from violent cosmic environments billions of light-years away. Their distribution seemed to follow a clean mathematical rule, simple, elegant, and stable. The newest IceCube analysis now suggests reality is far less orderly.
The most consequential industrial contests of the coming decade won't be visible from satellites. They won't be measured in gigawatts of installed capacity or kilometres of transmission line. They'll be measured in angstroms, in interface densities, in fabrication tolerances that determine whether a material converts ambient energy into electrical output or simply dissipates it as heat.