The universe is unfathomably vast, and for astronomers trying to understand it, that means gathering proportionately mind-boggling amounts of data. Wouldn’t it be great if there was something that could help speed up the search for patterns in all the trillions of galaxies out there and their four trillion stars?
The term “AI” has become a catchall these days for all sorts of dubious technologies of varying degrees of automation and reliability, but some types have found very practical and welcome uses among astronomers. For example, using custom-built AI tools, a team of European Space Agency scientists has identified more than a thousand “anomalies” in a collection of Hubble Space Telescope images that had gone unnoticed for decades. According to a NASA release.
Their work, described in A new study published in journal astronomy and astrophysicsThe first systematic search for astrophysical anomalies in the entire archive.
“Archival observations from the Hubble Space Telescope now span 35 years, offering a rich dataset in which astrophysical anomalies may be hidden,” said lead author David O’Ryan, ESA astrophysicist, in a NASA statement.
To make the discovery, the researchers used their AI tool, which they’re calling AnomalyMatch, on about 100 million snippets of Hubble images that were only a few pixels on each side. In less than three days, the neural network identified more than 1,300 anomalous objects, more than 800 of which had never been documented in the scientific literature.
The result is a veritable cosmic freak show. According to NASA, most of the anomalies were from galaxies colliding with each other, violent and highly disruptive events called galaxy mergers.
The AI tool has also identified a specific type of area called a jellyfish galaxyThey are defined by their numerous streams of star-forming gas that appear to hang from one side of the galaxy’s main disk, giving them the appearance of tentacles. Other oddities include edge-on planet-forming disks that look like hamburgers, and so-called gravitational lenses, in which the light of a giant foreground object like a galaxy bends the light behind it to act almost like a magnifying glass. And some discovered objects defied classification entirely, NASA said.
“This is a powerful demonstration of how AI can enhance the scientific recall of archival datasets,” Gomez said. “The discovery of so many anomalies not previously recorded in Hubble data underscores the instrument’s potential for future surveys.”
The clever automated anomaly hunter comes as NASA faces brutal cutbacks under the Trump administration, including the closure of its most historic facilities and even entire buildings, and mass layoffs. The administration has also eagerly deployed AI throughout the federal government Government-compliant versions of OpenAI modelsand an AI tool to help speed up the approval of drugs.
That said, astronomers have been working on AI solutions for some time. Commonly used to interpret large datasets like this latest work, it is also used to identify potentially habitable exoplanets and refine images of black holes. While the old guard in the field insist on caution when using these devices, it is clear that they have their uses.
More on space: These snapshots of the moment a star explodes will fill you with cosmic awe
