Li is certainly at the heart of the historical moment.
"I hope to bring AI technology to the most people and the most industries," she said, adding that this will have a profound impact on day-to-day living.
Known for AI research in computer vision, Li said technology in this area has become relatively mature, especially facial recognition and object tracking technology, which have been used in smart shopping, driverless cars and medical imaging and pathological analysis.
Potential application scenarios for AI are beyond count, Li said.
Financial services, media and entertainment, business, medical treatment, energy, education and manufacturing, among others, are deemed especially ripe for future growth.
"AI's promoting for the development of these industries has just started, but look at the massive demand."
What about entirely new industries that might be spawned through AI?
"When John von Neumann generated the idea for a computer, people never thought of software engineering as an industry," she said. "We need an adequate imagination."
SMART MACHINE VS SMART HUMAN
Commenting on the recent debate on whether AI will one day replace human intelligence, Li said the threat is overblown.
She cited a famous saying in the 1970s to clarify her view: The definition of today's AI is a computer that can make a perfect chess move while the room is on fire.
This means AI can accomplish plenty, such as memorizing 3,000 car models, but it doesn't understand the environment and context of every scenario, Li said.
"As a scientist, I want to keep the humbleness when I think about AI as a science as well as a technology," Li said. "So it's important to recognize AI as a very very young field. It still has a lot of open questions and challenges."
However, AI can play an important role in many areas with more human-machine cooperation, Li noted.
In cases such as the 2008 Wenchuan earthquake and 2011 Fukushima nuclear disaster, robots would have been more ideal to replace human beings in dangerous disaster relief work, she said.
Smart machines can also assist human beings in repetitive labor, Li said. For example, doctors can only diagnose a limited number of medical images, but machines can process a lot more at a likely lower cost and within a shorter time. This would allow doctors to conduct more valuable research and communicate more with patients, work that can't be replaced by AI.
"Machines don't have independent value," Li said. "The value of machines is the value of human beings."
