Renowned scientist Peter Ma has successfully harnessed the power of machine learning and artificial intelligence (AI) in processing data amassed by the Search for Extraterrestrial Intelligence (SETI) Institute, according to a recent press release.
Preliminary results hint at the possibility of discovering non-terrestrial “technosignatures”, potentially indicating the existence of extraterrestrial intelligence. This would mark a significant achievement, aligning with SETI’s ultimate objective.
In a groundbreaking research paper featured in Nature Astronomy, Ma elaborates on his strategy to train a machine-learning algorithm on 480 hours of telescope data from 820 stars compiled in 2016. Surprisingly, this AI method identified eight intriguing signals that were overlooked by earlier algorithms.
As an undergraduate at the University of Toronto, Ma expressed to VICE in an interview that their novel approach effectively eliminates the need for human intervention, setting it apart from previous machine learning techniques applied to SETI data.
Ma elaborated, “Our work is entirely reliant on the neural network without any support from conventional algorithms. This approach unearthed results that were undetected by traditional algorithms.”
The outcome of Ma’s innovative experiment is the identification of eight signals potentially originating from technologically advanced extraterrestrial civilizations. His algorithm specifically targeted signs that “are narrow band, doppler drifting signals originating from an extraterrestrial source.”
In his choice to employ a machine learning neural network, Ma recognized its adaptive nature, an advantage not found in more conventional AI algorithms. Given the uncertainty surrounding the characteristics of an extraterrestrial signal, Ma expressed the rationale behind their approach was simply to “learn it.”
Despite the detection of eight distinct signals, it remains unclear whether these signals indeed stem from alien societies.
The next phase entails further exploration of these signals in a bid to discern their origin and evaluate the need for subsequent observations in the corresponding regions of space.
Expanding on Ma’s pioneering work, the application of advanced AI techniques could revolutionize our approach towards the search for extraterrestrial life, offering a potential pathway to one of the most profound discoveries in human history. The interplay of machine learning and astronomical data could open new avenues in space exploration, and even if it doesn’t immediately lead to the discovery of alien life, these techniques could reveal fascinating insights about our universe.
