AI predicts that a dinosaur thought to be predatory may have been an herbivore

Researchers used artificial intelligence (AI) to discover information on dinosaurs, specifically to find out if a certain dinosaur was a predator or not. AI has shown that the dinosaur typically thought to be a carnivore may actually have been an herbivore.

The study was published recently in the Journal of The Royal Society Interface.

The new analysis was done by using deep convolutional neural networks (DCNNs) – computing systems based on neural networks of the brain – to track the fossils of these prehistoric reptiles. The study was conducted by Dr. Anthony Romilio, a paleontologist at the University of Queensland, and his teams from Australia, the U.K., and Germany.

The discovery

The team reanalyzed footprints from the Dinosaur Stampede National Monument in Central Queensland, Australia. For decades, scientists thought the giant footprints left by the dinosaurs were from a carnivorous dinosaur. “Large dinosaur footprints were first discovered back in the 1970s at a track site called the Dinosaur Stampede National Monument, and for many years they were believed to be left by a predatory dinosaur, like Australovenator, with legs nearly two meters long, said Dr. Romilio.

The discovery, using artificial intelligence, leads researchers to believe that the dinosaur was actually a friendly plant-eating prehistoric reptile. Dr. Romilio mentioned the difficulty in tracing the footprints from millions of years ago to determine the truth of what kind of dinosaur it really was. “The mysterious tracks were thought to be left during the mid-Cretaceous Period, around 93 million years ago,” he stated. But working out what dino species made the footprints exactly—especially from tens of millions of years ago—can be a pretty difficult and confusing business.”

That’s where AI comes in.

“So, to crack the case, we decided to employ an AI program called Deep Convolutional Neural Networks,” he continued. Also shortened as DCNNs, the AI program uses artificial neural networks to perform complete a detailed analysis of large amounts of data it is given through machine learning. The artificial intelligence computing system was trained with 1,500 dinosaur footprints, all of which consisted of animals that were theropod or ornithopod. These two groups were selected because they were the group of dinosaurs related to the Dinosaur Stampede National Monument prints.

The research and results

The outcome showed that the tracks had been made by an herbivorous ornithopod dinosaur. The dinosaurs started out as small and bipedal, standing on two legs, but were exclusively plant-eaters. The research team felt as if they were initially at a stalemate in determining the type of dinosaur that was at the site millions of years ago, and emphasized the huge significance of having AI to help them solve the mystery. “We were pretty stuck, so thank god for modern technology,” said Dr. Jens Lallensack, lead author from Liverpool John Moores University in the UK.

The research team consisted of three members who all had different prognostications on the type of dinosaur it was. “In our research team of three, one person was pro-meat-eater, one person was undecided, and one was pro-plant eater. So—to really check our science—we decided to go to five experts for clarification, plus use AI.”

The AI surpassed the researchers in declaring the dinosaur type. “The AI was the clear winner, outperforming all of the experts by a wide margin, with a margin of error of around 11 percent,” Dr. Lallensack said.

The researchers hope to utilize artificial intelligence in the future to further classify groups of dinosaurs, along with continuously building their database, using information they’ve discovered from the study.