A group of astrobiologists has unintentionally created an artificial intelligence (AI) system that has exceeded their expectations in identifying potential signs of alien life in materials recovered from space and other planets. The AI, developed and trained by the researchers, can determine with an impressive 90% accuracy whether samples are biological or nonbiological. What’s even more astonishing is that the AI, originally designed to classify samples into just two categories, autonomously expanded its capabilities to classify samples into three categories, including distinguishing between recently living organisms and ancient fossils. The researchers are intrigued by the AI’s unexpected capabilities, as they do not fully understand the inner workings of its identification process.
The AI has significant implications for the search for extraterrestrial life, being hailed as “the holy grail of astrobiology” by co-creator Robert Hazen. It offers a reliable and relatively simple method for analyzing samples, utilizing pyrolysis to break down materials into gas and decomposed parts for analysis. The AI’s ability to distinguish between different categories of biological samples, including fossils, opens up new avenues for exploring alien life and expanding our understanding of early life on Earth. This AI could potentially be used on robotic spacecraft, landers, and rovers to search for signs of life in space before returning samples to Earth, ushering in a new era in astrobiology and beyond.