Doomsday
A team of NASA astrophysicists has put the fate of entire star systems in the hands of an AI algorithm.
Called SPOCK by NASA and Princeton University astrophysicist Daniel Tamayo, the system doesn’t really decide which worlds will live and die. But it can predict the paths of exoplanets and determine which ones will remain stable and which will collide with other worlds or stars, much more accurately and on a larger scale than humans.
Too many files
Since the first exoplanet was discovered in 1995, scientists have identified more than 4,000 worlds elsewhere. More than 700 of them are in star systems containing more than one planet, Tamayo said in a press release, potentially putting them at risk of devastating collisions.
“We cannot say categorically ‘This system will be fine, but it will explode soon,'” Tamayo said in the statement. “The goal, instead, is, for a given system, to rule out all unstable possibilities that would have already collided and could not exist today.”
The reference
Traditionally, this is a problem that scientists would do with brute force by modeling the next billion orbits of an exoplanet and looking for danger. SPOCK is a bit more elegant: it stops after 10,000 orbits and then trains a machine learning algorithm using the dynamics of that orbit. Finally, the system learns to predict collisions well in advance.
“We call the model SPOCK – Stability of Klassifier Planetary Orbital Configurations – in part because the model determines whether the systems will ‘live and prosper,'” Tamayo said.
READ MORE: Artificial intelligence predicts which planetary systems will survive [Princeton University]
More about dead worlds: Scientists want to search for life on long-dead worlds
.
