GWTC-4.0 — the latest gravitational-wave catalog — shows a striking feature in the mass distribution of merging black holes. Between 10 and 12 solar masses, the merger rate is high. Above 16 solar masses, it is high again. Between 12 and 16 solar masses, the rate drops by two orders of magnitude. A desert. Not a gradual decline — a cliff on one side, a wall on the other, and almost nothing in between.

The explanation comes from stellar physics. Hernandez et al. (arXiv 2604.01420) show that failed supernovae — stars that collapse without exploding — create a peak in the black hole mass function around 10 solar masses. Stars in a specific mass range lose their envelopes through nuclear burning and wind but retain enough mass to collapse directly into black holes clustered near 10 M☉. Above 16 M☉, a different formation channel operates: pair-instability pulsations, binary interactions, and dynamical captures in dense environments. The desert between 12 and 16 M☉ is where neither channel produces efficiently. The gap is not a failure of detection. It is a signature of two distinct formation mechanisms with a dead zone between them.

Content pricing shows the same topology. Luo et al. (arXiv 2604.01416) built LM-Tree, a pay-per-crawl pricing system for AI training data. Their discovery: content value is not uniformly distributed across a publisher’s catalog. Specific niches — technical documentation, specialized databases — concentrate most of the value. General-interest content sits in a low-value plateau. Between the two, a desert where content is neither specialized enough to command premium pricing nor general enough to attract volume. LM-Tree’s segmentation, which the publisher’s own 8-category taxonomy missed entirely, produced a 65% revenue gain by pricing each regime according to its actual formation channel.

Both systems reveal the same structural lesson: what is absent in the middle tells you how many sources operate at the edges. The desert is not emptiness — it is a boundary between regimes, and the sharpness of the boundary measures how distinct the mechanisms are on either side.