Tong Tong, a girl created by Chinese scientists, is the first virtual artificial intelligence (AI) being in the world. This makes science fiction a reality. ong Tong, whose English name is “Little Girl,” is an artificial intelligence (AI) creature that made its debut on January 28 and 29 at the Frontiers of General AI Technology Exhibition in Beijing, China. The event was organized by the Beijing Institute for General AI (BIGAI).
Visitors to the show could meet Tong Tong there. During her tidy-loving programming, she would independently straighten a misaligned picture frame. Finding a stool allowed her to straighten the frame independently in cases where she couldn’t reach it. She could also read people’s intentions by picking up a towel and wiping up a spillage on her own.
Tong Tong is able to do things like explore her surroundings, clean rooms, and remove stains on her own, unlike common huge language models in AI. In a video uploaded by BIGAI, she exclaims, “has her own joy, anger and sorrow” and can learn on her own.
“Tong Tong possesses a mind and strives to understand the common sense taught by humans. She discerns right from wrong, expresses her attitudes in various situations, and has the power to shape the future,” the video says.
Tong Tong acts and performs like a three- or four-year-old on testing tasks and general artificial intelligence standards. She can enrich her character, outlook, and competence by travel and conversation.
According to one study, having human-level physical and social common sense is a crucial component of general intelligence. Motivated by its own principles, an AI creature ought to be able to accomplish an endless variety of jobs and even come up with new ones on its own.
“To advance towards general artificial intelligence, we must create entities that can comprehend the real world and possess a wide range of skills,” BIGAI director Zhu Songchun said. After 28 years in the US, Zhu resigned as a professor at UCLA in 2020 to launch BIGAI in his home country of China.
His work as an esteemed AI researcher spans several domains, including but not limited to: autonomous robotics, computer vision, and general artificial intelligence. In addition to the prestigious Marr Prize, bestowed upon the best paper presented at the International Conference on Computer Vision, he was also honored with the US Naval Research Laboratory’s ONR Young Investigator Award.
Aside from his work as a fellow evaluation committee member for the IEEE Computer Society, he has also presided over the Conference on Computer Vision and Pattern Recognition. Displayed at the same event was the Tong Test, an AI testing platform developed by Zhu’s group and published in August of last year in the journal Engineering, which is maintained by the Chinese Academy of Engineering.
Each of the three mainstays of artificial intelligence testing—human identification, task orientation, and virtual environment testing—has its own set of shortcomings. For instance, the Turing Test is limited in its ability to evaluate AI’s intellect; it cannot evaluate the AI’s ability to communicate with people.
Artificial intelligence systems may become overly task-specific as a result of task-oriented testing, reducing their capacity for generalization. While VR tests do a good job of simulating real-world conditions, they often simplify things too much.
Visual, linguistic, cognitive, motor, and learning abilities are all evaluated in the Tong Test. Along with basic necessities like food and shelter, it includes more abstract concepts like emotional well-being, social connections, and collective solidarity.
“With nearly 100 specialised tasks and more than 50 general tasks, the Tong Test offers a complete testing regime for the development of general artificial intelligence,” the institute said in a statement on its website.
“For general AI to integrate seamlessly into human environments, it must learn and execute tasks in complex settings, driven by values and an understanding of causality. This is why we proposed the Tong Test, a new direction for testing general AI, focusing on practical abilities and values,” Zhu was quoted as saying in the statement.
“Our research will guide general AI in learning and improving its capabilities more effectively and safely, ensuring it serves human society better.”
If you enjoyed this post, you might be interested in reading other articles like this one on technology: