Prehistoric hunter? No, says artificial intelligence.

AI has disrupted numerous industries and continues to accelerate financing, research, education, and technological advancements. AI research is exploding as AI algorithms get faster and less expensive to train. According to the AI Index Report, the quantity of AI research papers has increased by 300% between 1998 and 2018. Artificial intelligence is profoundly changing historical research, unearthing insights hidden in centuries-old records and enabling technological and historical analyses on a scale never before imagined. AI has also been used to decipher long-lost languages. Researchers at MIT's Computer Science and Artificial Intelligence Laboratory (CSAIL) have created a system that can decipher a lost language without prior knowledge of its relationship to known languages, which is a remarkable advancement given that previous research relied on mapping an unknown alphabet onto a known one. To connect texts to related terms in a known language, the system uses a decipherment algorithm based on historical linguistic principles.

Dr. Anthony Romilio of the University of Queensland employed AI pattern recognition in an international effort to re-analyze footprints from the Dinosaur Stampede National Monument, southwest of Winton in Central Queensland. "Large dinosaur footprints were initially discovered in the 1970s at a track site called the Dinosaur Stampede National Monument, and for many years they were thought to have been left by a carnivorous dinosaur, such as Australovenator, with legs nearly two meters long," Dr. Romilio explained. "The mystery tracks were considered to have been left some 93 million years ago during the mid-Cretaceous Period." However, determining which dinosaur species left the tracks, especially those dating back tens of millions of years, can be a challenging and perplexing task.

"This is especially true given that these large tracks are flanked by thousands of tiny dinosaur footprints, leading many to believe that this predatory beast may have provoked a stampede of smaller dinosaurs." So, to solve the situation, we opted to use Deep Convolutional Neural Networks, an AI program." It was trained using 1,500 dinosaur footprints, all of which were theropod or ornithopod in origin – the dinosaur groups relevant to the prints at the Dinosaur Stampede National Monument. The evidence was clear: the traces were created by a herbivorous ornithopod dinosaur. Dr. Jens Lallensack, the main author from Liverpool John Moores University in the United Kingdom, stated that the computer support was critical because the team was first stuck.

"We were fairly stuck," Dr. Lallensack remarked, "so thank God for contemporary technology." "One member of our three-person research team was pro-meat eater, one was uncertain, and one was pro-plant eater." So, in order to double-check our science, we decided to consult five specialists and employ AI. The AI was the clear winner, surpassing all of the experts by a considerable margin, with an error margin of roughly 11%. "When we applied the AI to the big tracks from the Dinosaur Stampede National Monument, all but one of these tracks were firmly recognized as having been left by an ornithopod dinosaur – our prehistoric predator." The team aims to continue expanding the fossil dinosaur track database and conducting additional AI research.

The study was published in the Journal of the Royal Society Interface and involved researchers from Australia, Germany, and the United Kingdom. The Queensland Museum in Brisbane has a replica of the dinosaur trackway on exhibit, and the track site may be seen in Winton, Queensland.