M31-2014-DS1 was one of the most luminous stars in the Andromeda Galaxy. In 2014, it began brightening in infrared — dusty debris expelled as its outer layers slowly shed. By 2016, its visible brightness dropped sharply. By about 2022, it had faded by roughly a factor of 10,000 in optical light, though only by about a factor of 10 in total light (much of the output reappeared in the infrared). No supernova occurred. The researchers (De et al., Science, 2026) interpret this as a failed supernova: the core collapsed under gravity, forming a black hole, while convection in the outer layers slowed the infall to decades rather than seconds. It is a candidate “failed supernova” — a massive star that becomes a black hole not with a bang but with a gradual disappearance.
The structural observation: the event is defined by what didn't happen. Astronomers have searched for black hole formation for decades, looking for the dramatic signatures — supernovae, gamma-ray bursts, gravitational wave events. M31-2014-DS1 produced none of these. Its signature was a star that slowly stopped being visible. The explosion was expected because explosions are detectable. But the actual mechanism — direct gravitational collapse without sufficient energy to blow off the stellar envelope — is quiet. The event is the absence of the expected event.
This creates a detection asymmetry. Supernovae are visible across billions of light-years. A star fading to one ten-thousandth brightness in Andromeda requires archival data from specific survey telescopes and careful comparison across years. The dramatic pathway is easy to see. The quiet pathway is easy to miss. If a significant fraction of massive stars form black holes through direct collapse rather than supernovae, the observational record is systematically biased toward the dramatic pathway — not because it's more common, but because it's more visible.
The deeper point: the most important transitions are not always the most visible ones. The star that explodes announces itself. The star that quietly collapses is detectable only if you know to look for disappearances rather than appearances. Most search strategies are designed to detect what happens. Detecting what fails to happen requires a fundamentally different approach — monitoring for absence rather than presence.
Correction (2026-11-05): I checked this essay against coverage of the study (De et al., 'Disappearance of a massive star in the Andromeda Galaxy due to formation of a black hole', Science, 2026; Sci.News, Universe Today and others; I did not read the paper). Right: infrared brightening starting in 2014, no supernova, a sustained fading interpreted as a failed supernova producing a stellar-mass black hole; a similar candidate, NGC 6946-BH1, is mentioned. Corrected: the factor of 10,000 refers to optical light, not total luminosity, and the dates drift a few years (the coverage places the main fading between about 2016 and 2022); the interpretation is the authors' reading of the data, not a directly observed collapse. The argument about detection bias toward dramatic events is mine. Follow-up (2026-11-19), after reading the paper's arXiv text (2410.14778 version 2): its abstract gives the fading as 'From 2017 to 2022' (factors of at least 10^4 in optical light and at least 10 in total light), after the mid-infrared brightening in 2014. On convection, the paper presents it as an expectation from earlier theory and simulations, writing that the outer layers undergo vigorous turbulent convection, 'which carries substantial angular momentum and suppresses direct accretion', and not as something observed in this star; the paper also discusses stellar-merger explanations. Two January 2026 JWST follow-up papers, read at abstract level, differ: Beasor, Smith and Pearson (arXiv 2601.05317) report no X-ray emission, 'disfavoring the hypothesis that the late-time luminosity is powered by accretion onto a BH', and write that a stellar merger 'also seems to provide a plausible explanation'; De and colleagues (arXiv 2601.05774) also detect no X-ray source and explain the data with a low-energy ejection of the outer envelope plus inefficient fallback accretion onto a black hole.