Liu Qingfeng, Chief Executive Of The University Of Technology, Inc.: The Cash Dividend Of The Application Of Artificial Intelligence Is Coming.

2020/1/3

Gan Jun photo

In 1956, the concept of artificial intelligence was first put forward, but until the ten years after 2010, with the progress of big data, cloud computing and Internet of things, the development of AI has won massive data base and extensive application scenarios.

Looking forward to the next ten years, the wheels of new and old energy conversion will move forward. As one of the representative technologies of the new round of scientific and technological revolution, AI will undoubtedly play an important role.

So, when the wheel passes over the current year, what track did the past year leave in the history of AI development? "In 2019, AI was ushered in scale applications, and the cash dividend period is coming." Recently, in an exclusive interview with the twenty-first Century economic report, Liu Qingfeng, chairman of the University of information technology, said.

What Liu Qingfeng said is not just for nothing. In his view, whether or not the cash dividend of artificial intelligence is coming is mainly based on three criteria: first, whether there are real application cases that can be seen and touched in the market; secondly, whether these real cases can be generalized; in addition, we need to explore whether the case can prove its effectiveness by using unified data, and how much value it will play in many aspects such as cost effectiveness and labor substitution.

"Through these three rules, we can see that the bonus of artificial intelligence has indeed arrived." Liu Qingfeng said.

Industry ushering in spring

According to the "new generation AI plan" published by the State Council, by the year 2030, the scale of the downstream core industry of AI will reach trillion yuan, and the scale of related industries will reach 10 trillion yuan. Today, the application of AI has begun to accelerate in many industries.

"In the 2019, in the field of education, artificial intelligence has made it impossible for students to teach students in accordance with their aptitude for thousands of years." Liu Qingfeng pointed out.

According to the new generation of artificial intelligence development research center of the Ministry of science and technology and the Roland Begg management consulting company released the "intelligent education innovation application development report", from 2014 to 2016, it is the three highest year of the intelligent education industry. In the three years, 43 intelligent education companies were newly established, accounting for 56% of the total number of 2011-2018 years of intelligent education companies.

However, the education industry under the support of intelligent technology has undergone tremendous changes. According to Liu Qingfeng, in the course of preparing lessons for teachers, intelligent education products can save material saving time by 53%, and machine automatic correction can save 42% of the time. At the same time, the efficiency of classroom teaching has increased by 26%. And students can get personalized learning push when they get home. Through targeted exercises, the rate of error solving is as high as 80%.

In addition to education, with the continuous progress of AI technology, the development of many industries has ushered in disruptive changes. For medical industry, artificial intelligence will play an important role in health care. It will become one of the main trends in the field of health care in the future. Precision medical treatment and cost reduction demand will become the main driving factor of AI in the medical industry.

Liu Qingfeng said that in the application scenario of hospitals, the related artificial intelligence products can cover the whole process service before, during and after diagnosis. Through the medical history collection, voice electronic medical record, mobile medical care system, follow-up after diagnosis, we can achieve accurate, complete collection, real-time and effective presentation of patients' medical data, comprehensively improve the professional level and work efficiency of medical personnel, improve the patient's experience and improve the hospital service ability and quality.

In addition, artificial intelligence is also playing a different role in translation, automobile, city, finance and other fields. "Over the past three years, we have always stressed that artificial intelligence is a technology of empowerment, and its core value is to help our industry improve its effectiveness." Zhang Hongjiang, director of Research Institute of Chi Yuan, told the twenty-first Century economic news reporter: "AI can enable cars to drive automatically, and use computer vision technology to identify faces faster and more accurately. AI has also made a series of progress in financial fraud prevention and credit reporting.

Technical bottleneck is not equal to cold winter.

At the same time, the core of the current wave of artificial intelligence is deep learning algorithm.

At the NeurIPS Conference on artificial intelligence held this month, Blaise Agueray Arcas, Google's top researcher, pointed out that deep learning technology can not solve all problems, especially tasks related to reasoning and social intelligence.

Yoshua Bengio, director of the Mila Institute of artificial intelligence, also said that the deep learning model is highly specialized and therefore very narrow. "They need more samples than humans to learn a task, but even so, they will make a very stupid mistake."

However, Liu Qingfeng always believed that the deep explanation of the algorithm and its limitations were not enough to show that artificial intelligence was in the cold winter. "During the third wave of AI, the scale landing of related applications and the resulting data information can solve the problem of application effectiveness." Liu Qingfeng stressed that "during this period, we can promote products that can effectively improve people's work and life efficiency, and the third wave will not be shattered."

Liu Qingfeng believes that the future breakthroughs of artificial intelligence will be reflected in three aspects. First of all, the basic theory of small sample, unsupervised and personalized problems brings about a continuous breakthrough in the algorithm. Secondly, the combination of brain science research and mathematical statistical modeling method forms the combination of brain and intelligence. In the end, it is a continuous exploration of artificial intelligence system and human behavior cooperation.

It should be noted that the development of 5G technology will also become the "combustion supporting device" of artificial intelligence. "5G boosts the tide of interconnection of all kinds of things. Voice interaction will dominate the mode of human-computer interaction." Liu Qingfeng pointed out that "in the era of Internet Interconnection, more and more devices will be used in the absence of screen, mobile and far-field conditions. As the most natural and convenient way of communication, speech will become a vital entry for all devices."

This also means that small to smart watches, large to mobile cockpit - intelligent terminal devices will usher in the outbreak of the 5G era. At the same time, 5G technology will also bring about the explosion of content, such as high-definition video, games, and so on, and will also increase the chance of AI application.

"For example, for content processing, automatic editing of content, or adding some voice tagging to content, there will be new opportunities for development in these areas." Liu Qingfeng analysis said.

Artificial intelligence management year

With the rapid development of the industry, the shadow behind the technology has also appeared unprecedentedly, triggering a series of discussions about technology boundaries, management standards and privacy protection.

In August 2019, a APP called "ZAO" suddenly fires all over the net. Users only need to upload a photo in APP, so that they can replace their faces into big stars. The effect is almost false. But soon, this APP is controversial because of alleged infringement of user privacy. Users upload face information, will be faced with infringement, theft and other security risks.

Based on the related APP threats to user privacy and the risk of data leakage, the state Internet Information Office, the Ministry of culture and tourism and the State Administration of radio and television jointly issued the "Regulations on the management of network audio and video information services", which have been implemented since January 1, 2020, and restrict the production, release and dissemination of audio and video information based on new technologies such as deep learning and virtual reality.

In addition, a series of artificial intelligence problems, including data monopoly, algorithm discrimination, intelligence abuse, data poisoning, ethics and so on, have attracted more and more attention.

"Artificial intelligence legislation to protect human life and property safety, that is to say, AI must be used in the controllable range of risk, for example, in the field of automatic driving, its application threshold and standard is what; in the medical field, artificial intelligence must achieve certain accuracy, and it is better to pass certain qualification authentication." Liu Qingfeng stressed that "only those standards and conditions that meet these criteria can be used in the field of practical use."

Besides the need to keep pace with legislation, head enterprises also need stronger protection consciousness. "Now there are many AI services related to the daily life of ordinary people, so leading enterprises in related industries should be exemplary, not only to inform users, but also to guide values." Liu Qingfeng pointed out.

At the same time, he said, at present many data are taken away by the enterprise without the user's knowledge. "From a technical point of view, we should give users more choices through the development of AI technology."

http://www.sjfzxm.com/global/en/559724.html