Artificial Intelligence , Now Aiding In the Search for , Extraterrestrial Intelligence.<br />According to a new paper published in the journal <br />'Nature Astronomy,' scientists are hoping to apply <br />machine learning to the search for extraterrestrial life.<br />SETI, or the Search for Extraterrestrial Intelligence, was <br />established in 1984 and has been scanning space for radio <br />signals comprised of non-Earth based "technosignatures.".<br />VICE reports that a machine-learning <br />algorithm was applied to telescope <br />data collected in 2016. .<br />After analyzing over 480 hours of data from <br />820 stars, the algorithm identified eight signals <br />of interest that had previously gone undetected.<br />Peter Ma, the paper's first author, said that while <br />artificial intelligence has been applied to SETI's data <br />in the past, his team's approach is something new.<br />Previously people have <br />inserted ML [machine learning] <br />components into various pipelines <br />to help with the search. , Peter Ma, Undergraduate student at <br />the University of Toronto, via VICE .<br />This work relies entirely on just the neural <br />network without any traditional algorithms <br />supporting it and produced results that <br />traditional algorithms did not pick up, Peter Ma, Undergraduate student at <br />the University of Toronto, via VICE .<br />According to Ma, his team's work is twice as fast as <br />traditional algorithms and allows for an out-of-the-box <br />approach to the search for alien intelligence.<br />Traditional algorithms operate on <br />a given set of instructions designed <br />by us… thus the algorithm will only <br />ever discover what we tell it to find, Peter Ma, Undergraduate student at <br />the University of Toronto, via VICE .<br />The issue is that the nature of an ET signal <br />is not completely known… Hence our <br />proposed approach is to just learn it, Peter Ma, Undergraduate student at <br />the University of Toronto, via VICE
