by Xinhua writers Zhang Chunxiao, Xu Xingtang, Qin Lang
REDMOND, the United States, Dec. 3 (Xinhua) -- The world will continue to see technological breakthroughs in artificial intelligence (AI), and their potential application in healthcare and financial services will have a transformative impact on human life, Microsoft's AI chief has said.
Currently, the application of AI in medicine is limited to the reading of medical images, such as an electrocardiogram, which is in essence image recognition, Harry Shum, executive vice president of Microsoft's AI and Research group, told Xinhua in a recent interview.
In the future, he said, breakthroughs will take place in the intersection of AI and genetics, "which is so-called precision medicine."
Shum envisioned that the integration of AI, big data, and genetic analysis of sensor data will help improve the understanding of human body.
"I believe such breakthroughs are definitely going to happen, perhaps very soon, which will be truly incredible," he said.
Greater use of AI breakthroughs is also expected in financial services, which would change many of today's conventional practices, Shum said, citing AI and big data used in high-frequency trading.
If AI technologies are widely applied in finance, the whole market could be drastically transformed, he said.
The Chinese computer scientist, who joined Microsoft in 1996, is responsible for driving the company's overall AI strategy and forward-looking research and development efforts spanning infrastructure, services, apps and agents. He oversees AI-focused product groups including Bing and Xiaoice.
Recent years have seen major AI advances in a number of areas, with the most prominent taking place in gaming and perception, according to Shum.
Citing Google's AlphaGo, Microsoft's mahjong AI, and Pluribus, a poker bot designed by Facebook and Carnegie Mellon University, he said these games have grown to be highly intelligent through AI deep learning and reinforcement learning.
AI has come a long way in perception, particularly in voice recognition, voice synthesis and visual design, Shum said, predicting that in five to 10 years, AI could "defeat" humans in this regard.
However, he added, when it comes to cognition, the progress is less impressive. "Compared with human intelligence, AI is still in a quite early stage in terms of its understanding of the world, things and people around, and in particular, itself."
Today, AI has already permeated all kinds of industry, Shum said, warning of "many serious problems" that extensive application of AI could bring about to society.
For instance, he said, self-driving vehicles are one of the most exciting cases of AI application in transportation. With governments, big companies and venture capital putting in a huge amount of money, full self-driving will eventually become a reality.
The biggest problem facing self-driving, however, is not a technical one, he said. It might take about five to 10 more years to solve all technical problems.
"The greatest challenge is beyond technology. For example, the most dangerous moment is when half of the vehicles on the road are driven by people and the other half self-driving," he said.
To smooth the road to self-driving, rules and regulations are yet to be rolled out, legal issues to be addressed, and ethical questions to be answered, such as how a self-driving car should be designed to react in times of traffic accidents, he noted. "It might take many more years to tackle these issues."
In addition, Shum said, AI advances could aggravate the unfair distribution of wealth.
"In the future, technical, social and administrative intervention measures will have to be taken to address this serious issue," he said.