AI development will lead to a bright future for HK

It might be a bit of an exaggeration to say that artificial intelligence is the most exciting technology in the world today. But the development of this technology is now occurring very rapidly - with every advance taking the world by surprise.

Last year, South Korea was astonished by AlphaGo's defeating legendary Go player Lee Se-dol, a ninth-dan professional with 18 world titles. This led to the South Korean government appropriating additional funds for a joint effort to build a national AI research center with several conglomerates, including Samsung, LG Electronics and Hyundai Motor. As a result, the budget for the research center was increased by 55 percent to a total of 1 trillion won ($866 million).

When discussing a country's ambitions with AI, perhaps China is No 1 in this regard. In July, the State Council rolled out the "Next-Generation Artificial Intelligence Development Plan". It is aiming to catch up with the forerunners in AI technology development and applications by 2020, and become a dominant artificial intelligence innovation power in the world by 2030. With such ambitions in mind, the Chinese authorities have boosted training skilled people with considerable urgency. Great efforts are being made to produce people with the right skills. This includes the introduction of AI-related courses in primary and secondary schools, promoting the study of coding and setting up more AI courses in tertiary education.

According to a study published by the McKinsey Global Institute in April, the number of academic papers Chinese researchers have published so far in reputable AI publications was less than half of that of US researchers. But China has become increasingly influential in this domain. A renowned academic organization - Association for the Advancement of Artificial Intelligence (AAAI) -re-scheduled its 2017 annual meeting just to avoid clashing with the Chinese New Year in order to encourage more Chinese scholars to attend the meeting. In another major academic conference on the AI field, the 26th International Joint Conference on Artificial Intelligence (IJCAI) held in Melbourne, Australia, in August. It attracted over 2,000 AI researchers from all over the world. Professor Yang Qiang, head of the Department of Computer Science and Engineering of the Hong Kong University of Science and Technology, was elected chairman of the Awards Assessment Committee. Yang was the first Chinese to assume such an important role since the establishment of this conference. In 2013, he was also elected as the first Chinese fellow of AAAI, which only grants a few scientists such an honor each year. This suggests the mainland and Hong Kong's scientific research capacity in AI is very strong and is up to date.

Unfortunately, AI research and development, albeit flourishing, basically is still confined to universities and international companies only. Most of the research and development results have not yet been widely applied by local industries. Therefore, I am delighted that the Hong Kong Observatory has tackled the challenges of processing terabyte of data with AI technology such as machine learning and deep learning. The self-developed Intelligent Meteorological Monitoring Assistant filters the daily weather data, equivalent to 150 million pages, with the built-in 200 rules, and generates useful summaries, forecasts and warnings. Riding on the AI technology, the Observatory not only provides more accurate weather information to the public, it also won the Gold Prize of Team Award (Internal Service) in the Civil Service Outstanding Service Award Scheme 2017.

If more public and private organizations make greater efforts to adopt new AI technology as the Observatory did, we can be sure that significant improvements will be made to our quality of life. Road congestion, for example, is one of the most common problems in cities - including Hong Kong. According to a study by the Texas A&M Transportation Institute, traffic congestion caused urban Americans to travel an extra 6.9 billion hours and purchase an extra 3.1 billion gallons of fuel with a total cost of $160 billion in 2014. When Adaptive Signal Control Technology has been introduced in the US to allow traffic lights to change based on AI technology and real time data, travel time in general has been shortened by more than 10 percent.

In China, Alibaba Cloud (AliCloud) has worked with the Hangzhou government to provide a smart traffic management service called Hangzhou City Brain. Its primary goal is to alleviate traffic congestion through video and image recognition technologies with images from about 50,000 roadside surveillance cameras. The green light, for example, can be automatically extended when the system detects a vehicle coming, helping to shorten waiting times. As a result, congestion has been reduced and the speed of the traffic is increased by as much as 11 percent in areas where AliCloud can collect sufficient data.

According to the latest QS study, Hong Kong has four universities ranked among the top 50 in the world (South Korea and Japan have only two, respectively, while the Chinese mainland has three). Hong Kong has a big pool of top-notch AI researchers and a strong capacity for scientific research. If we can make better use of AI, we can not only enhance our quality of life and help retain skilled people, but also inject a new impetus into the local economy and provide more room for young people to unleash their creative potential.

(HK Edition 10/03/2017 page5)