Text|perseverance
Source|Bowang Finance
The concept of artificial intelligence has ushered in a positive situation again.
On July 13, the Cyberspace Administration of China, together with the National Development and Reform Commission, the Ministry of Education, the Ministry of Science and Technology, the Ministry of Industry and Information Technology, the Ministry of Public Security, and the State Administration of Radio and Television, promulgated the Interim Measures for the Management of Generative Artificial Intelligence Services, which clearly stipulates the development and governance, service specifications, supervision and inspection, and legal responsibilities of generative AI technology, which will come into force on August 15, 2023.
Source: Cyberspace Administration of China official website.
The official launch of the above measures reflects the attitude of regulators to the development and risk of artificial intelligence at this stage, and will initially establish a regulatory system for the field of generative artificial intelligence in the mainland.
Affected by this, A-share artificial intelligence stocks rebounded comprehensively on the first trading day after the release, and related concept stocks such as AI computing power, servers and information security rose sharply, and some of them directly rose to the limit.
How long can artificial intelligence still be popular?
01
Under the intensive introduction of favorable policies, the artificial intelligence industry is promising
Since it was selected as one of the "2017 Chinese Media Top Ten Buzzwords" in December 2017, artificial intelligence has attracted much attention from all walks of life, and since this year, it has set off a boom driven by the intelligent chat tool ChatGPT, from the global volume to China, and from the technology industry to all walks of life.
In order to promote and regulate the development of artificial intelligence, the mainland has intensively introduced favorable policies. The Politburo meeting at the end of April pointed out that it is necessary to attach importance to the development of general artificial intelligence, create an innovative environment, and attach importance to risk prevention; In May, the State-owned Assets Supervision and Administration Commission also guided and urged central enterprises to increase their layout in strategic emerging industries such as new-generation information technology and artificial intelligence.
In addition to the top-level design of the central government, local governments have also responded positively and proposed measures and methods with their own characteristics, further refining the development path of artificial intelligence.
To give a few examples, Beijing issued Several Measures to Promote the Innovation and Development of General Artificial Intelligence; Shenzhen will accelerate the high-quality development and high-level application of artificial intelligence, and build a city-level intelligent computing power platform; Shanghai will provide up to 20 million yuan of support for key AI technology projects; Chengdu proposed 19 financial support measures, with the highest award of 10 million, to support and encourage the development of the AI industry.
The mainland's artificial intelligence industry is also beginning to take shape. According to the calculation results of the China Academy of Information and Communications Technology, the scale of the core artificial intelligence industry in mainland China reached 508 billion yuan by the end of 2022, a year-on-year increase of 18%; The number of enterprises is 4,227, accounting for about 16% of the total number of enterprises in the world. A number of AI open platforms such as iFLYTEK, Baidu, Huawei, Tencent, and Alibaba initially have the ability to support the rapid development of the industry.
Of course, the popularity of artificial intelligence will not fade in the short term.
According to the relevant person in charge of the Ministry of Industry and Information Technology, "during the '14th Five-Year Plan' period, the mainland will accelerate the expansion of the artificial intelligence industry and cultivate a good development ecology", and specific measures include: organizing an innovation consortium jointly participated by large and small enterprises and industry, academia and research, promoting breakthroughs in key core technologies of artificial intelligence, and improving the level of mainland smart chips, development frameworks, and typical smart products; Accelerate the application of artificial intelligence in manufacturing, transportation, energy and other fields, and promote intelligent transformation in key areas; Build industrial clusters, cultivate a group of internationally competitive artificial intelligence leading enterprises, develop a number of specialized and special new enterprises, and rely on the pilot area to build an industrial agglomeration development highland.
It is predicted that by 2030, the scale of the core artificial intelligence industry in the mainland is expected to exceed 1 trillion yuan, promote labor productivity by more than 20%, and drive the scale of related industries to exceed 10 trillion yuan.
It can be said that the mainland artificial intelligence industry has broad prospects, and in the process of deep integration with the real economy, it has also become a new growth engine for economic and social development.
02
The new regulations were "released" and model governance was proposed for the first time
Three months after the release of the draft for comments, the new rules for generative artificial intelligence have been freshly "baked", which can more effectively promote the healthy development of generative AI.
In this regard, various brokerages have also given positive responses. For example, Orient Securities said in the research report that "the implementation of the Measures has initially established a regulatory system for the field of generative artificial intelligence in the mainland, and it is expected that the commercialization process from basic models to upper-level applications in the subsequent AI field will be rapidly advanced, and the AI ecology is expected to be rapidly established." GF Securities also pointed out that "independent innovation, safe and credible AI basic technology is encouraged by policies and is expected to develop rapidly".
The new regulations put forward that the state adheres to the principle of attaching equal importance to development and security, promoting innovation and governance according to law, takes effective measures to encourage the innovative development of generative artificial intelligence, implements inclusive and prudent and categorical supervision of generative AI services, and clarifies the overall requirements for the provision and use of generative AI services. In addition, it also stipulates security assessment, algorithm filing, complaints and reports, etc., and clarifies legal responsibilities.
The most epoch-making significance of this new regulation is the first proposal of model governance and the establishment of a regulatory system covering the life cycle of generative artificial intelligence.
This also means that technology governance has entered a new stage of development.
First of all, in the definition, generative artificial intelligence technology refers to the model and related technologies with text, pictures, audio, video and other content generation capabilities, the model is not only the technical engineering before the application of generative artificial intelligence, but also directly affects the final generated content, it integrates technical supporters, service providers and content producers at the technical level, so it is necessary to specifically take it as a governance object.
Secondly, in terms of management, the model is the gripper of generative artificial intelligence service supervision, and through training, it has become a new governance object that integrates data, algorithms, and computing power. Article 4 of the New Regulations, "The provision and use of generative AI services shall comply with laws and administrative regulations, and respect social morality and ethics", which clarifies the provisions to be complied with, including: taking effective measures to prevent discrimination in ethnicity, belief, country, region, gender, age, occupation, health, etc. in the process of algorithm design, training data selection, model generation and optimization, and providing services; Respect intellectual property rights and business ethics, keep trade secrets, and must not use the advantages of algorithms, data, platforms, etc. to carry out monopoly and unfair competition behaviors; Respect the lawful rights and interests of others, must not endanger the physical and mental health of others, and must not infringe on others' portrait rights, reputation rights, honor rights, privacy rights, and personal information rights and interests.
Finally, the concept of "basic model" is innovatively proposed, and Article 7 "Generative AI service providers shall carry out training data processing activities such as pre-training and optimization training in accordance with law" must first comply with "using data and basic models with legal sources".
In addition, the new regulations also have several highlights, such as optimizing the adjustment responsibility mechanism, accurately adjusting the scope of application, and improving training data requirements.
03
Artificial intelligence concept stocks rose in response to the trend, and enterprises accelerated their nuggets
Affected by the favorable policies of the new regulations, artificial intelligence concept stocks have risen in response to the situation.
According to the "Economic Reference Network", in response to the introduction of the new regulations, Zhou Hongyi, founder of 360 Group, said, "The content shows the attitude of relevant departments to encourage innovation and support the development of large models, and provides confidence and guarantee for the innovation of science and technology enterprises", while pointing out that "the applicable boundary is clarified, which can not only standardize the public service of large models, but also facilitate the research and development and application of organizations." At present, industrial digitalization has become a national strategy, and the development of large models should also empower industrial digitalization, as an industrial revolution-level productivity tool to help traditional industries, especially manufacturing, achieve digital transformation to intelligent upgrading. The real future opportunity for the big model is the enterprise-level market, including organizations, governments, and cities. Judging from the Measures, the state has a strong attitude of strong support for the application research and development and implementation of enterprise-level scenarios."
As a leading enterprise in the domestic large model industry, 360 proposed the "two wings flying together" artificial intelligence big model development strategy, while developing core technologies, combined with its own scenario advantages, seized the opportunity of large model commercialization and scenario-based landing, and has launched enterprise-level AI large model solutions, and took the lead in landing applications in taxation, enterprise services and other industries, achieving a significant increase in productivity.
Let's look at iFLYTEK, with the halo of "the first company listed on the A-share market with artificial intelligence as its main business", and also joined the big model track. In early May this year, iFLYTEK officially released and displayed its large-model product "Spark Cognitive Big Model", which is applied in education, office, automotive, digital employees and other fields. According to reports, all technical modules of iFLYTEK Xinghuo cognitive large model are completely independent research and development, whether it is underlying training, external service reasoning, or data, modeling, reinforcement learning, there is a complete innovation system, which is the key to iFLYTEK being able to give a clear time node.
In response to the question of "whether the new regulations meet expectations", iFLYTEK replied on July 14 that "the company participated in the discussion of multiple rounds of revised draft management measures." In addition, the iFLYTEK Spark large model will be upgraded as scheduled on August 15 as originally planned, when the large model version will break through the code capability and upgrade the multi-modal interaction capability."
But it is worth mentioning that last year iFLYTEK ended the myth of "10 consecutive rises". According to the 2022 financial report, iFLYTEK's revenue increased slightly to RMB18.82 billion that year, but the net profit attributable to the parent fell sharply by 64% year-on-year to RMB561 million, which was also the first time in nearly five years that the net profit declined.
In fact, from the perspective of subdivided business, the artificial intelligence industry also presents differences, among which the field of optical modules has made rapid progress.
Zhongji Xuchuang, the leader of A-share optical modules, said, "With the release of the generative artificial intelligence big language model represented by ChatGPT, the surge in demand for AI computing power has been generated, which in turn has driven a significant increase in the demand for communication products including optical modules, and accelerated the iteration of optical modules to 800G and above products."
There is no doubt that although domestic AI governance is at the forefront of the world after the introduction of the new regulations on generative AI, the mainland artificial intelligence industry still has problems such as a relatively weak industrial foundation such as high-end chips and a slight shortage of high-level talents, which still need to be actively faced.