If a radio telescope ever intercepts something genuinely artificial from beyond Earth, the hardest problem won’t necessarily be receiving it. It will be recognizing it, deciding fast whether the pattern buried in the noise is a real message or nothing at all, and that’s part of what John Elliott presented at IAUS404: Advancing the Search for Technosignatures in March 2026.
About the Presenter
Dr. John Elliott is an Honorary Research Fellow in the School of Computer Science at the University of St Andrews and coordinator of the university’s SETI Post-Detection Hub.
He has worked on SETI post-detection problems since the late 1990s, publishing on signal decipherment, computational linguistics, and dolphin communication as models for recognizing non-human intelligence. In 2012, alongside Astronomer Royal Martin Rees, he co-founded the UK SETI Research Network, which he still chairs.
Arthur C. Clarke once praised his work personally, and it has been cited multiple times in NASA’s 2018 technosignatures report.
More About Elliott’s Work in Advancing the Search for Technosignatures
Elliott’s talk, subtitled “information, affinity, autonomy, and design,” tackled a problem we face after a potential detection: once a receiver has captured a candidate signal, how do you tell whether it carries information at all? His approach borrows from decades of his own research comparing structures across human languages, dolphin whistles, and even artificial languages robots have invented for themselves, looking for mathematical “universals” that might mark out intelligence regardless of species. He described building an “affinity matrix” that measures how closely an unknown signal’s structure resembles known communicative systems, paired with entropy calculators and AI reasoning engines to automate the analysis at a speed no human could match.
He also proposed a “Situational Reasoning Compass” for guiding an AI’s response depending on context, and revisited the DISC Quotient, a post-detection response framework he co-developed with science-fiction author Stephen Baxter. Since the talk, Elliott has published a peer-reviewed paper expanding this entire framework for autonomous, uncrewed probes that may have to interpret the unknown without waiting for instructions from Earth.
Key Takeaways
Recognizing a technosignature may be less about the antenna and more about the analysis: filtering structured information out of noise is where AI becomes essential.
Elliott’s “affinity matrix” measures how closely an unknown signal resembles known communication systems, human, animal, or even AI-generated, as a first step toward classification.
The DISC Quotient, developed with Stephen Baxter, offers a graded framework for how humanity might calibrate its response based on a signal’s apparent content.
Because interstellar distances may rule out real-time guidance from Earth, Elliott argues these same recognition and decision tools need to be built directly into autonomous probes.
For a reader who wants to go further, Elliott’s May 2026 open-access paper in Frontiers in Astronomy and Space Sciences, “Machine learning for discovering life beyond Earth,” lays out the full architecture in far more technical detail than the talk allowed.
This video was included in the proceedings of the International Astronomical Union (IAU) symposium #404, Advancing the Search for Technosignatures, hosted by Blue Marble Space.
Thanks for joining us here with Planetary Perspectives, an editorial endeavor and science communication project from Blue Marble Space.
