New AI model claims Milky Way’s black hole spins at “top speed” — but experts warn of iffy data
An international team used AI to analyze noisy data from the Event Horizon Telescope (EHT) and created a new image of Sagittarius A*, the supermassive black hole at our galaxy’s center. Their AI-enhanced image suggests the black hole spins nearly at top speed, and its rotational axis points straight at Earth.
The EHT links telescopes worldwide but struggles with atmospheric interference, making its data extremely tricky to interpret. The team trained a neural network on previously discarded, noisy telescope data — data too messy for traditional methods. 
Michael Janssen, astrophysicist at Radboud University and study co-author, told Live Science the AI handles the tough EHT data well: "A neural network is ideally suited to solve this problem."
The results dropped this month in Astronomy & Astrophysics. Pinpointing the black hole’s spin could reveal key details about how matter behaves near it.
But the findings aren’t airtight. Reinhard Genzel, Nobel laureate and astrophysicist at the Max Planck Institute, urged caution.
Reinhard Genzel stated:
"I’m very sympathetic and interested in what they’re doing,"
"But artificial intelligence is not a miracle cure."
He worries the low-quality data fed into the AI may have biased the model, producing a distorted image that shouldn’t be taken at face value.
Janssen’s team plans to test their AI on fresh EHT data soon to see how well it holds up against real-world measurements.
Image credit: EHT Collaboration/Janssen et al.
Source: LiveScience.com
