AI Deciphers Ancient Dinosaur Footprints

An international team of researchers has developed an AI-powered application called Dinotracker, designed to identify dinosaur footprints from the ancient past. This innovative tool analyzes thousands of fossil and simulated footprints, focusing on eight key characteristics like toe width and heel position to classify the tracks.

The algorithm was trained to be unbiased, providing a neutral mathematical perspective to assist human experts. When tested, the AI achieved a 90% agreement rate with human classifications, demonstrating its effectiveness in identifying dinosaur species from their tracks.

Dinotracker's capabilities have already led to intriguing possibilities regarding bird evolution. The AI detected strong similarities between ancient footprints and bird foot structures, suggesting birds may have evolved earlier than previously thought.

However, researchers emphasize that skeletal remains are the true evidence for evolutionary timelines. The AI is available for free on GitHub, requiring some technical knowledge to implement, and its data pool is expected to grow as more experts use it.