AI Shows that this Dinosaur May Have Been a Friendly Herbivore

Not all dinosaurs are carnivores because one of them may have been a vegetarian, according to artificial intelligence.

A group of researchers employed AI to learn more about dinosaurs, especially if they were a predator or not, and they recently found out that one dinosaur that is commonly regarded to be a predator may have actually been a friendly herbivore, as reported first by Interesting Engineering.

Tracking Dinosaur Footprints

Deep convolutional neural networks (DCNNs), which are computer systems based on brain neural networks, were used in the new investigation to examine the fossils of these ancient reptiles. Paleontologist Dr. Anthony Romilio from the University of Queensland and his team from Germany, the UK, and Australia carried out the study.

The team reexamined footprints in the Dinosaur Stampede National Monument in Central Queensland, Australia.

According to Dr. Romilio, for many years it was thought that large dinosaur footprints found in the 1970s at the Dinosaur Stampede National Monument were left by a carnivorous dinosaur with approximately two-meter-long legs like Australovenator.

The team opted to use Deep Convolutional Neural Networks, an AI tool, to solve the issue. The AI software, also known as DCNNs, employs artificial neural networks to do an exhaustive study of the substantial volumes of data that machine learning has provided.

The 1,500 dinosaur footprints used to train the artificial intelligence computer system were all from theropod or ornithopod species.

These two dinosaur species were chosen because they shared a connection with the prints discovered at the Dinosaur Stampede National Monument, as noted by Interesting Engineering.

AI Finds a Herbivore

The findings indicated that a herbivorous ornithopod dinosaur had left the tracks. The earliest dinosaurs were small, bipedal, stood on two legs, and only consumed plants.

The research team stressed the critical importance of having AI to assist them in solving the mystery because they first felt at a standstill in identifying the sort of dinosaur that was at the site millions of years ago.

Three people made up the research team, and each of them had a different theory about what kind of dinosaur it was.

One individual supported eating meat, one was unsure, and one supported eating plants. The team opted to apply AI in addition to consulting five specialists to thoroughly verify its science.

In identifying the dinosaur kind, the AI outperformed the researchers. According to Dr. Lallensack, the AI model surpassed all of the experts by a large margin and had an error margin of about 11%.

The researchers intend to use the knowledge they have gained from the study to continue building their database and use AI to further categorize groups of dinosaurs.

This article is owned by Tech Times

Written by Jace Dela Cruz