Welcome to DU! The truly grassroots left-of-center political community where regular people, not algorithms, drive the discussions and set the standards. Join the community: Create a free account Support DU (and get rid of ads!): Become a Star Member Latest Breaking News Editorials & Other Articles General Discussion The DU Lounge All Forums Issue Forums Culture Forums Alliance Forums Region Forums Support Forums Help & Search

Judi Lynn

(164,027 posts)
Mon Dec 1, 2025, 01:27 AM 21 hrs ago

A.I. Reveals Signs of Early Life in 3.3-Billion-Year-Old Rocks. Next, It Could Continue the Search in Space

The new approach looks at the distribution of molecular fragments in material, allowing for broad surveys in degraded specimens

Margherita Bassi - Daily Correspondent
November 26, 2025

Paleobiologists investigating Earth’s earliest life traditionally study tiny fossils or hardened structures made by ancient microbes. The oldest ones date to around 3.5 billion years ago. But the structures are exceedingly rare. Most were absorbed and shattered by Earth’s crust before churning back up to the surface.

Another method searches for biomolecules, chemical compounds related to living things, trapped in rocks. This approach for finding direct evidence of life has reliably identified these molecules in rocks up to 1.7 billion years old. But such rocks also get chewed up by Earth’s ever-changing crust, destroying many traces of old organisms.

Now, researchers have demonstrated an artificial intelligence-based technique that can aid the search for ancient life—and it doesn’t require the presence of entire molecules. The approach, described in a paper published November 17 in the journal Proceedings of the National Academy of Sciences, pieces together evidence of life from the distribution of biomolecular fragments.

Using this method, researchers identified chemical fingerprints of life in rocks more than 3.3 billion years old, and they uncovered evidence that plants started producing oxygen through photosynthesis almost one billion years earlier than previously documented. The team says their new A.I. technique could even aid the search for extraterrestrial life.

“We can tease out whispers of ancient life from highly degraded molecules,” Robert Hazen, a mineralogist and astrobiologist at the Carnegie Institution for Science and co-author of the study, tells Reuters’ Will Dunham. “This is a paradigm shift in the way we look for ancient life.”

More:
https://www.smithsonianmag.com/smart-news/ai-reveals-signs-of-life-in-3-billion-year-old-rocks-next-it-could-continue-the-search-in-space-180987715/

Latest Discussions»Culture Forums»Science»A.I. Reveals Signs of Ear...