Personal profile:
Jiao Licheng is a Professor in the School of Electronic Engineering at Xidian University, Xi'an, China. He is the Dean of the Electronic Engineering School and the Director of the Key Lab of Intelligent Perception and Image Understanding for the Ministry of Education of China at Xidian University. Dr. Jiao is a Senior Member of IEEE. His research focuses on image processing, natural computation, machine learning, and intelligent information processing. He has published more than 20 monographs and 500 papers in international journals and at conferences.
Since 2017, the focus of China's Artificial Intelligence (AI) policy has shifted from pure research to industrial convergence. The development of AI depends on the innovation of theories and algorithms as well as the upgrade and update of computing chips. Traditional computing architectures do not support large-scale parallel computing of AI algorithms. In the future, a computing revolution will be required to accelerate AI computing processes. The focus of AI chip research has changed from GPUs and FPGAs to TPUs, and from non-customization to customization.
In 2006, Geoffrey Hinton and his student Ruslan Salakhutdinov proposed a deep learning model, which marked the beginning for a third wave of AI. Thanks to the popularity of the Internet, the explosive growth of data, and the development of key information technologies, AI has been applied in a wide array of fields. The United States, Japan, South Korea, and other countries have all developed their own strategic development plans for AI. In 2017, the State Council of China issued the Next Generation Artificial Intelligence Development Plan, which is a clear timetable and roadmap for the development of AI in China. Now, China regards AI as a national strategy. Just as mankind moved from an agricultural revolution to an industrial revolution and then to an information revolution, the time for AI+ has arrived.
The explosive growth of data and the significant improvement of processing capabilities have been fundamental in making the leap from the Big Data era to the AI era. AI is data-driven, cross-field, and convergent. AI would not be possible but for a mix of factors. First is the development of Big Data algorithms. The rapid update of various AI algorithms can excavate the principles behind data and tap into the value of Big Data. A second factor is the relationship developing between humans and our interactions with intelligent terminals. With the rapid development of the Internet, the Internet of Things (IoT), and intelligent terminals, the interactions between people, people and machines, and machines and machines is increasingly frequent. Now, multimedia information, such as text, audio, still images, and videos, is transmitted dynamically. AI is improving multimedia information processing technologies and has made breakthroughs in speech recognition and image recognition.
China is fully embracing AI technology, and has intensively introduced an array of AI promotion policies following the May 2015 release of the China Manufacturing 2025 initiative. The goal is for public and private companies to leverage AI so they can profit from its benefits. These policies focus on AI technologies in emerging fields such as smart phones, shopping, and security.
Mature technologies, such as fingerprint, facial, and image recognition, machine translation, and speech recognition and synthesis will all become deeply integrated in the future. The integration will enable human-like intelligent applications. For now, these technologies at the intelligent computing and decision-making layers need to be further developed. Although still in the early stages, technologies at the decision-making layer, for example, autonomous cars and humanoid robots, can completely revolutionize our way of life.
Outlook on AI Applications
AI can improve the capabilities and explore the potentials of many industries and professions. AI+ — the mass application of AI technology throughout industry — has become a hot topic that is affecting finance, manufacturing, transportation, healthcare, and education.
Intelligent robots include home service robots, industrial robots, and medical robots. Home service robots can be family companions. Industrial robots can replace human beings in dangerous work environments. Medical robots can help doctors improve the accuracy of surgical operations and assist patients in recovery.
According to industry research, the market for intelligent driving is expected to reach approximately USD $19.4 billion (CNY 121.4 billion) by 2020. Computer vision and deep learning are core technologies for intelligent driving. The convergence of multiple assisted driving technologies will be adapted to more scenarios and even enable unmanned driving. The full automation of intelligent driving will significantly improve the security of road transportation and allow a wider range of people to enjoy convenient travel.
The finance industry has transitioned from financial office digitalization to Internet finance, and now to AI-based finance. AI is being integrated into people's economic life through mobile payments, digital currency, intelligent investment consultation, risk control, and mobile banking are all playing an important role in optimizing traditional financial structures. Laying these foundations now means that future scholars and industries can easily further explore the dividends brought by combining AI and finance.
In the education sector, intelligent tools based on AI are being developed, such as an intelligent scoring system, intelligent assessment system, and intelligent learning system. The results will help to slash the cost of teaching and learning equipment and improve the learning efficiency of students. AI algorithms can be used to capture and analyze students' learning modes and enable teachers to customize teaching plans for specific students.
In the medical sector, intelligent software provides health management for patients and users. Intelligent devices will one day replace doctors for specific work. Image recognition and speech recognition have been used to assist doctors to improve the accuracy of diagnostics and treatments that improve the working efficiency of medical institutions and personnel while reducing cost.
Moving Forward Decisively
Around the world, the rapid development of AI has brought admirable achievements. However, many basic theories of AI need to be perfected. In recent years, the deep learning algorithm has become the focus of AI development and has made great breakthroughs in speech recognition, image recognition, and machine translation. Oriented to problems and data, deep learning still has a long way to go in its research, but despite this, a variety of research principles and methodologies are still emerging and will continue to emerge in the future.
The development of brain and cognitive science in the early 20th century initially inspired the research of AI. Coupled with a greater understanding of brain and cognitive science, AI will be leveraged to eventually help machines achieve or surpass human beings' intelligence. Brain-like intelligence and hybrid intelligence are the latest research directions of AI.
The development of AI depends on the innovation of theories and algorithms as well as the upgrade and update of computing chips. The traditional computing architecture cannot support the large-scale parallel computing of AI algorithms. In the future, a computing revolution will be required to accelerate the computing process. The focus of AI computing chip research has changed from GPU, FPGA, to TPU and from non-customization to customization. Companies such as NVIDIA, Google, and Huawei are actively developing computing chips that can better adapt to AI development. Additionally, quantum computing will bring revolutionary development opportunities for AI. The number of quantum bits increases exponentially. Miniaturized quantum chips make possible the rapid and real-time processing.
China's AI achieves major breakthroughs in basic theories, technologies, and applications every five years. By 2030, China is projected to lead the world's AI development and become the world's leading AI innovation center. Another key element to drive the development of AI lies with the highly-skilled human workforce. The AI and Automatic Driving Vehicle Research Report launched by the Aminer team in Tsinghua University indicates that there are more AI scholars in colleges than those in institutions and enterprises. By the end of 2017, more than 50 universities had set up AI majors. However, current undergraduates, graduates, and PhDs who specialize in AI are far from meeting the needs of future industry development. Therefore, the pace of cultivating a highly-proficient AI workforce needs to be accelerated.
The third wave of AI is in full swing, and basic research and technical achievements are the keys to surviving the fierce competition.
