The Kepler space telescope, which launched in 2009, has produced more than 30,000 signals measuring light from stars to search for possible planets in distant parts of the galaxy. Sifting through that and other telescope data, astronomers have found more than 3,500 planets, up from 329 known before the Kepler mission. Now artificial intelligence is helping to find even more.
On Thursday, researchers from Google and University of Texas at Austin announced that a machine-learning algorithm had discovered two new planets, Kepler 80g and Kepler 90i. Kepler 90i is a particularly special discovery—it’s the eighth planet orbiting its star, marking the first system outside our own known to have eight planets.
To find the planets, the researchers trained an algorithm on 15,000 labeled signals from Kepler data, provided by NASA, to recognize what is and isn’t a planet. In tests, the algorithm could correctly identify which signals reflected planets and which did not 96 percent of the time.
After that successful test, the researchers focused on searching through signals around 670 stars that were known to have planets already, which is where they found two new planets that they’d missed before.
“It’s like needles in a haystack,” Chris Shallue, a senior engineer at Google AI and one of the researchers on the project, said in a conference call about the discovery.
It’s a good thing researchers are now armed with AI to help in their search, because there’s plenty more data to get through. During its four-year initial mission, Kepler observed 200,000 stars, creating about 14 billion data points in the process. Shallue said that with so much data, their algorithm stands to find a lot more planets—including ones with Earth-like conditions.
This artist is dominating AI-generated art. And he’s not happy about it.
Greg Rutkowski is a more popular prompt than Picasso.
What does GPT-3 “know” about me?
Large language models are trained on troves of personal data hoovered from the internet. So I wanted to know: What does it have on me?
An AI that can design new proteins could help unlock new cures and materials
The machine-learning tool could help researchers discover entirely new proteins not yet known to science.
DeepMind’s new chatbot uses Google searches plus humans to give better answers
The lab trained a chatbot to learn from human feedback and search the internet for information to support its claims.
Get the latest updates from
MIT Technology Review
Discover special offers, top stories, upcoming events, and more.