In the modern age, the search for extraterrestrial life requires a unique mind. A researcher must have the openness and creativity to imagine something beyond our current knowledge of the universe. At the same time, the researcher, if they are to be taken seriously, must also be skilled enough to analyze high-powered astronomical imagery through supercomputing, artificial intelligence, or other means.
The researchers behind project Hephaistos, a Swedish-based effort to identify traces of alien life in the cosmos, fit this bill. They are extensively published astrophysicists with a shared dream: that the observable universe may contain signs of far-away civilizations yet unrecognized.
Earlier this year, the group published a paper documenting their use of machine learning to comb through large datasets of infrared images for Dyson spheres — networks of alien satellites harvesting energy from a star. The idea was old, but the scientists’ AI-forward approach was new.
“If you want to do this, you kind of have to try something that hasn’t been tried before,” says project Hephaistos leader Erik Zackrisson.
In their paper, project Hephaistos identified seven candidates for further research — star’s that appeared to be encircled by something like a Dyson sphere.
