Fast radio bursts are among the strangest signals in astronomy: millisecond flashes of radio energy, some releasing in a thousandth of a second what the Sun emits in days, arriving from across the cosmos. By 2020 the field had a leading story for what makes them. They come from magnetars — young, violently magnetized neutron stars, the still-warm corpses of massive stars that recently died in core-collapse supernovae. That model got a huge boost when a magnetar inside our own galaxy was caught firing an FRB-like burst. Young, fresh, explosive: that was the profile. Then a burst named FRB 20200120E was pinned down to its home, and the home was an old-age home.

The localization is the evidence, and it is unusually good because this source is close. FRB 20200120E was found to repeat, and a global network of radio telescopes using very-long-baseline interferometry — work led by Franz Kirsten and colleagues, published in Nature in 2022 — pinpointed it to roughly the M81 galactic system, only about 3.6 megaparsecs away, making it the nearest extragalactic FRB known at the time, around forty times closer than any other. But the precise position is the shock. The burst does not come from M81's star-forming disk. It sits within a globular cluster, only about two parsecs from that cluster's center.

That single fact is the anomaly, and it is hard to overstate. Globular clusters are the fossil neighborhoods of the universe. They are tight, ancient swarms of hundreds of thousands of stars, and their populations are old — on the order of 9 to 12 billion years, with effectively no recent star formation. There are no massive young stars left to die. There has been no core-collapse supernova there for billions of years. So the favored FRB engine — a magnetar born hours or years ago in the explosion of a giant star — cannot be operating in this cluster. The signal is coming from exactly the place where it should be impossible by the textbook model.

This is why the result mattered more than just adding one dot to a catalog. It is a clean falsification of universality: whatever else is true, not all fast radio bursts come from young magnetars, because at least one demonstrably does not. The astronomers behind the localization proposed the alternatives that an old environment allows. You can still make a highly magnetized neutron star late in cosmic time without a fresh supernova — for instance through the accretion-induced collapse of a white dwarf that has been fed past its limit, or through the merger of two compact stellar remnants. And globular clusters are famous engines for exactly those compact binaries, because their dense cores let stars swap partners and pile into tight pairs far more efficiently than the thin galactic disk does.

The skeptical-but-fair reading is careful here, and it cuts against sensationalism. Nothing about FRB 20200120E requires new physics, aliens, or exotic sources. It requires the field to admit that 'fast radio bursts' is probably a category with more than one cause — the same way 'things that go bang in the sky' once lumped together novae, supernovae, and gamma-ray bursts before they were sorted out. The anomaly is not that the laws of nature broke in this globular cluster. It is that our tidy single-origin model was too tidy, and one well-measured burst from the wrong neighborhood was enough to break it.

What stays open is the actual machinery. We now have a strong case that this particular source is an old, recycled compact object rather than a baby magnetar — but 'accretion-induced collapse' and 'compact-object merger remnant' are still hypotheses competing over a single nearby example. The unresolved question is whether FRB 20200120E is a rare exception or the visible tip of a whole second population of bursts powered by the dead and the ancient rather than the newly born — a population we have been quietly miscounting because we assumed every flash had to come from something young.