GoKawiilNew AI model OneStrata recasts three of Earth's 'Big Five' extinctions as gradual declines
Researchers trained a machine-learning model, OneStrata, on global fossil-record data rather than text, using it to reconstruct biodiversity changes across hundreds of millions of years. The model, presented at the Geological Society of America meeting in Denver on October 12, indicates that three of the so-called Big Five mass extinctions unfolded as gradual declines rather than sudden catastrophic events.
GoKawiil's interpretation of the reporting above, not reported fact.
If confirmed, the finding could prompt palaeontologists to revise long-held narratives about sudden biological catastrophes, since Shuzhong Shen of Nanjing University calls the approach a 'real revolution' for studying past diversity change. However, Matthew Clapham of UC Santa Cruz suggests the model may add detail without fundamentally overturning existing high-resolution local reconstructions, indicating the finding's true significance remains debated among specialists.
- OneStrata applies language-model-style pattern recognition to fossil data instead of text.
- The model suggests three of the 'Big Five' mass extinctions were gradual, not sudden.
- Experts are divided on whether this reshapes or merely supplements existing extinction research.
Source: nature.com — Dinneen, 2026-10-20
Published there as: “Vast new fossil record challenges story of Earth’s ‘big five’ mass extinctions”
Read the original report → The summary and analysis above are GoKawiil's own, written from reporting by the source above. Facts and quotes belong to the original publisher.