The search for extraterrestrial intelligence (SETI) consists of looking for evidence of extraterrestrial intelligence originating beyond Earth, trying to detect indications of advanced non-human technology, which could have been developed by alien civilizations. The most common technique is searching for radio signals. Radio is an excellent medium for sending information across the incredible distances between the stars; It zips through the dust and gas that often lie between stars and Earth, and it does so at the speed of light—about 20,000 times faster than our best rockets. Many SETI projects use giant satellite dishes to pick up radio signals that hypothetical extraterrestrials might transmit.
When thinking about the probability of discovering technologically advanced extraterrestrial life, the question that often arises is: “if they are out there, why haven’t we found them yet?” And often the answer is that we have searched only a tiny part of the galaxy. Furthermore, algorithms developed decades ago for early computers can be outdated and ineffective when applied to modern petabyte-scale data sets. And the trend is that the amount of data to be scrutinized continues to increase by gigantic steps, taking into account that technological innovations in the field of the search for extraterrestrial intelligence (SETI) allow scrutiny of the cosmos on a scale every time. elderly.
Now, the team of Peter Ma, from the University of Toronto in Canada, and Cherry Ng, from the SETI Institute in the United States, has applied a deep learning technique (a modality of artificial intelligence) to a previously studied data set of stars. nearby and has uncovered eight intriguing signs that were once overlooked.
Antennas like this can be used to communicate with distant spacecraft or to scan the cosmos for artificial signals of non-human origin. (Photo: NASA JPL/Caltech)
In total, the new AI system has examined 150TB of data from 820 nearby stars, in a dataset that had previously been examined in 2017 using classical analytic techniques and mislabeled as lacking in interesting signals.
Artificial intelligence can review data much faster than a human and find hints of extrahuman artificial signals that are too subtle for a human to notice directly.
These qualities of artificial intelligence will be increasingly necessary to analyze the data obtained in the SETI surveys, especially if we take into account that initiatives that will examine a million stars are already underway. As Ma argues, the use of artificial intelligence systems like the one her team has used can dramatically speed up the rate of suspicious signal discovery.
Specifically, Ma and his colleagues used artificial intelligence to reanalyze data that was collected by the Green Bank radio telescope in West Virginia, United States, as part of a Breakthrough Listen campaign that apparently found nothing of interest.
The scrutiny made by the artificial intelligence system, and additional checks carried out by the human members of the team to confirm what was found, have revealed receptions of eight signals that, due to the peculiarity of their characteristics, could be artificial (although for now there are no way to check).
The signals were narrowband, that is, they had a narrow spectral width, on the order of a few hertz. Signals caused by natural phenomena tend to be broadband.
The signals appeared in observations oriented to a specific point in the sky and not in observations away from that point. If a signal comes from a specific celestial source, it appears when we point our instrument towards the target and disappears when we move it away from that point. Radio interference of man-made origin usually occurs both when pointing at and away from a celestial target, due to the proximity of the source.
Other of the examined traits are also rare in signals of natural origin.
Ma and his colleagues discuss the technical details of their AI data analysis in the academic journal Nature Astronomy, under the title “A deep-learning search for technosignatures from 820 nearby stars.” (Fountain: NCYT de Amazings)