On November 14, the 26th China International Hi-Tech Fair (hereinafter referred to as the Hi-Tech Fair) with the theme of "Technology Leads the Development of Industrial Integration and Fusion" kicked off at the Shenzhen International Convention and Exhibition Center.
This year's Hi-Tech Fair has more than 5,000 high-tech companies from more than 100 countries and regions participating in the exhibition, including more than 180 of the world's top 500 companies and state-owned enterprises, more than 1,000 listed companies, "gazelle companies" and "unicorn companies" There are more than 2,000 companies, and the professional audience is expected to reach 500,000.
At the "China High-tech Forum" of this year's Hi-Tech Fair, academicians of the Chinese Academy of Engineering, presidents of multinational companies, well-known entrepreneurs and other industry figures shared their exploration of science and technology and cutting-edge technologies around the theme of "New Productivity Leads Future Industrial Development" think.
"2024 China High-tech Forum" on-site picture source: provided by the organizer
At the forum, Chen Qingquan, academician of the Chinese Academy of Engineering, director of the Electric Vehicle Research Center of the Hong Kong Polytechnic University and distinguished chair professor of the Department of Electrical and Electronic Engineering, said: "Now all industries are improving new quality productivity...New quality productivity needs to get rid of traditional economic growth. The method generally has the characteristics of "new three highs", that is, high technology, high efficiency and high quality. The "old three highs" model of high investment, high consumption and high emissions is replaced by the "new three highs".
In Chen Qingquan’s view, the past agricultural economic era was dominated by individual labor, and the industrial economic era was dominated by large-scale production. But now we are entering the digital economic era, which will regard computing power as its core competitiveness.
"I am mainly engaged in research in the field of electric vehicles. The biggest insight that the electric vehicle industry has given me is that automobile innovation cannot be carried out in isolation. What I proposed is the 'four network integration', that is, through the transportation network, energy network, information network, and humanities network Integration can generate endless value opportunities in various industries,” said Chen Qingquan.
Gao Wen, academician of the Chinese Academy of Engineering and director of Pengcheng Laboratory, focused on this year's Nobel Prize winning situation, discussed the deep impact of artificial intelligence on society, and mentioned the importance of relevant competitions in identifying and rewarding innovation.
"In the era of intelligence, the scale of computing power is the core element of international competitiveness...The advantage of computing power does not lie in the volume of computing power, but in higher computing efficiency, extensive emerging technology applications and sound infrastructure support." Gao Wen express.
He said that an important turning point in the artificial intelligence industry occurred in 2012. "At that time, the industry generally used its own artificial intelligence models to participate in competitions. At that time, the competition standard was whoever made the system with a lower error rate would win. In 2010, the team with the lowest error rate was 28%, and in 2011 it was 26%. , achieved a breakthrough in 2012, falling to 16%. After 2013, the deep neural network technology used by the team that achieved the breakthrough began to become popular. Later, the technology was applied to automatic machine recognition and automatic sorting systems, including today’s speech recognition, which is basically based on neural networks. The technical route of the Internet.”
"(Promoting) the accelerated development of the artificial intelligence industry, scientific competitions are very important. In addition, large models of artificial intelligence must have their own base and their own computing power, so that they can carry the intelligence and inheritance we want to achieve." Gao Wen said.
The rapid development of the artificial intelligence industry and its practical applications in various industries are topics frequently mentioned by many industry figures on the forum. However, with the rapid development of the industry, more and more attention has been paid to the security risks and hidden dangers emerging in the artificial intelligence industry.
"At this stage, the artificial intelligence industry is turbulent, and technological innovation is changing with each passing day. Even last year, there was a 'battle of hundreds of models' in China. The development of the industry has now entered a new stage. Not only do we feel the gushing impact of artificial intelligence, With the development speed, more experts are also paying attention to the safety hazards and risks of artificial intelligence during this development process," shared Xie Qinglin, member of the party group and deputy general manager of China Electronics Information Industry Group Co., Ltd.
Xie Qinglin said: "Artificial intelligence needs to be more safe. For example, with the large models that are used more often now, I guess everyone has had the experience that the output results are not so reliable, and sometimes people even comment that 'it is talking nonsense seriously'. This is actually It’s a relatively small point, but if we don’t have enough grasp of the infrastructure, underlying technology, etc., once an extreme situation occurs, as small as a large model, as large as a system or even the entire industry, we may face more security threats. and challenges.”
In Xie Qinglin’s view, the security of artificial intelligence requires both the security of intelligent computing and the security of “key roots.”
"Judging from the practical experience of China Electronics, there are three main directions to solve the security problems of intelligent computing, namely system security, platform security and supply chain security. These three security are the key and core of what we call intelligent computing security. ." Xie Qinglin said.
Xie Qinglin said that as the current financial, telecommunications, energy and other industries have increasingly strong demands for intelligent computing, and at the same time have put forward higher requirements for security and reliability, the advantages of the independent computing industry chain will become more and more significant. .
In addition, when talking about some of the problems currently plaguing artificial intelligence companies, Yang Zhen, CMO, head of strategy and marketing of Nortel Digital Intelligence, said that many companies basically fall into three categories: "dare not", "unable" and "unwilling". dimensions to consider.
“The so-called ‘dare not’ refers to doubts about domestic computing power, uncertainty about the performance of domestic computing power, and worries about compliance and security pressure; ‘no’ means mainly from business-level considerations, even if the computing power can "Satisfaction" refers to whether it can support actual business after use; "unwillingness" mainly includes unwillingness to reconstruct product business and unwillingness to invest funds and manpower. "Yang Zhen believes that most companies currently focus on exploring the field of artificial intelligence. the above three aspects.