Global Industry Trends Weekly: What future investment opportunities does the World Artificial Intelligence Conference showcase?

Recent market trading hotspots mainly focus on structural investment opportunities in new energy industries such as N-type modules, industrial and commercial energy storage, and lithium battery materials. In terms of industry trends, the 2023 World Artificial Intelligence Conference will be held at the Shanghai World Expo Center and Expo Exhibition and Convention Center from July 6 to 8, and it is recommended to focus on investment opportunities in the fields of robotics, autonomous driving, large models and data elements. The policy side mainly focuses on the platform economy, digital economy, artificial intelligence, new energy vehicles and medicine.

Core ideas

【Market trading hotspot】The recent market trading hotspot mainly revolves around structural investment opportunities in the new energy industry. Hot spot 1: N-type modules, upstream polysilicon prices stop falling to promote photovoltaic installed demand, subdivided attention to N-type module investment opportunities. Hot spot two: industrial and commercial energy storage, cell prices are reduced, superimposed peak-valley price difference is expanded, and the economy of industrial and commercial energy storage is highlighted. Hot spot three: lithium battery materials, new energy vehicle sales exceeded expectations, driving the demand for lithium battery materials.

【Global Industry Trends】Focus on structured investment opportunities under the 2023 World Artificial Intelligence Conference. The 2023 World Artificial Intelligence Conference, held from July 6 to 8 at the Shanghai World Expo Center and Expo Exhibition & Convention Center, recommends focusing on investment opportunities in the areas of robotics, autonomous driving, large models and data elements. Robotics: It is recommended to pay attention to the investment opportunities of linear joints with roller screws as the core and rotary joints with precision reducers as the core; Autonomous driving: It is recommended to pay attention to investment opportunities related to ADAS and lidar; Large models: Continue to pay attention to investment opportunities related to the implementation of domestic large models such as Huawei Pangu 3.0 and their scenario applications, and data elements: Follow up investment opportunities in high-quality data supply, data componentization, and data annotation.

【Important Policy Tracking】Platform Economy: Zhejiang issued the Implementation Opinions on Promoting the High-quality Development of the Platform Economy, and the current focus of platform economy policies has shifted to "supporting healthy development"; Digital economy: Beijing issued the Implementation Opinions on Better Leveraging the Role of Data Elements to Further Accelerate the Development of the Digital Economy, which is the Beijing version of the "20 Data Articles", and data elements are accelerating; Artificial intelligence: Shanghai issued Several Measures to Promote the Innovation and Development of Artificial Intelligence Large Model in Shanghai, which plans to build three plans and five platforms around four major directions: innovation capabilities, innovation elements, innovative applications, and innovation environment; New energy vehicles: The Ministry of Industry and Information Technology and other relevant departments have revised the Measures for the Parallel Management of Average Fuel Consumption and New Energy Vehicle Credits for Passenger Vehicle Enterprises, which put forward higher requirements for the innovation and development of new energy vehicles; Medicine: The National Health Insurance Administration issued a draft of renewal rules for comments, which is expected to stabilize corporate confidence and help normalize the renewal of contracts in the country.

【Important Information】Huawei: HUAWEI CLOUD Pangu Model 3.0 officially released. Pangu Model 3.0 is a completely industry-oriented large model, which is divided into three layers: L0 basic large model, L1 industry large model, and L2 scene model, which will reshape thousands of industries. In addition, the research results of HUAWEI CLOUD's Pangu meteorological model were published in the official journal of Nature, which is the first paper published by a Chinese technology company as the only signatory in recent years. Toyota: On July 4, Toyota said it has made a major breakthrough in solid-state battery technology, which can halve the weight, volume and cost of batteries. Keiji Kaita, president of Toyota's carbon neutral research and development center, said that he believes that it is now possible to create a solid-state battery with a range of 1200 kilometers and a charging time of 10 minutes or less.

【Short-term Industry Trend Concerns and Suggestions】From a short-term perspective, July needs to focus on five major tracks with marginal improvement: integrated circuits (storage, packaging and testing, materials), digital economy (data elements), new energy vehicles (complete vehicles, batteries, autonomous driving), AI+ (applications, algorithms, computing power), and photovoltaics (N-type modules).

【Medium and Long-term Industry Trends Concerns and Suggestions】From a medium- and long-term perspective, we recommend paying attention to the rise of the whole industry chain of carbon neutrality (photovoltaic, wind power, energy storage, hydrogen energy, etc.) under the global attention, the increase in the penetration rate of electric smart vehicles under the general trend of electrification and intelligence, and industrial trends such as digital economy, artificial intelligence, metaverse, and autonomous controllability.

【Global Stock Market Performance】Global stock markets rose and fell last week. In terms of industries, the thematic indices with higher A-share gains last week were: air transport select, automobile vehicle selection, and new energy vehicles. In terms of sectors, the global stock market, public utilities and optional consumer sectors performed relatively well last week, while the healthcare and information technology sectors performed relatively poorly, and among the companies with a market value of US$20 billion that rose sharply in the past week, most of them were companies in the energy and optional consumer sectors; Among the $20 billion companies with market value that led the decline in the past week, most of them were companies in the information technology and health care sectors.

【Risk Warning】The individual stocks mentioned in this article only indicate that they are related to the relevant theme and do not constitute investment advice.

Contents

01 Core Concerns and Investment Advice

1. Preview of important matters this week

2. Hot spots in market transactions

Last week, the top A-share gainers of the important theme indexes were air transport selection, automobile vehicle selection, and new energy vehicles.

Note: The stocks involved in this article are listed or represent related to the relevant industrial chain, do not represent individual stock investment suggestions, nor do they represent the research views of China Merchants Industry.

Last week and this week's market trading hotspots are structural investment opportunities in new energy, mainly focusing on the following three directions:

? N-type components

The decline in upstream polysilicon prices has promoted PV installed demand, and the subdivision focuses on investment opportunities in N-type modules. On July 5, according to the data released by the Silicon Industry Branch, the average price of compound feeding materials was 67,500 yuan / ton, the average price of dense materials was 65,700 yuan / ton, unchanged from the previous month, and the average price of N-type materials was 74,200 yuan / ton, up 0.68% from the previous month. According to the view of China Merchants New Group, polysilicon prices continued to fall, head companies returned to single-digit profit margins, and second-tier losses were widespread, basically the price bottom. At present, the polysilicon price can support a large shipment of 1.4 yuan/W modules, and after the completion of destocking next month, the subsequent terminal demand is expected to have strong and positive changes sooner. At present, the premium of domestic N-type TOPCon modules and batteries to current products continues to climb, and the profit advantage of the N-type industrial chain is obvious and continues to expand.

• Industrial and commercial energy storage

The price of battery cells has decreased, and the peak-valley price difference has expanded, and the economy of industrial and commercial energy storage has become prominent. Since 2022Q4, the prices of lithium carbonate and lithium iron phosphate, which account for the highest proportion of battery costs, have been reduced, and the pressure on the battery cells at the energy storage cost end has been eased. With the general expansion of the national peak-valley price difference, the number of provinces and cities with the national peak-valley price difference exceeding 0.7 yuan/kWh has increased to 8; Coupled with Zhejiang, Anhui, Guangdong, Hainan and other provinces and cities, the implementation of the two-charge and dual-discharge strategy has effectively improved the utilization rate of the energy storage system and shortened the cost recovery cycle. In addition, the increase in photovoltaic installed demand will also benefit the investment of energy storage enterprises. Segmented areas, focusing on energy storage PCS investment opportunities.

? Lithium battery materials

Sales of new energy vehicles exceeded expectations, driving demand for lithium battery materials. The association expects retail sales of NEVs in June to be about 670,000 units, up 15.5% month-on-month and 26.0% y/y, with a penetration rate of about 36.6%. In terms of related manufacturers, on July 1, a number of new energy vehicle companies disclosed their sales in June, most of which achieved all year-on-year growth. Ideal delivered 32,575 units in June, up 150% year-on-year and 15.2% month-on-month; BYD's June sales of new energy vehicles reached 253,000 units, a year-on-year increase of 88.79%, and the cumulative sales volume for the whole year increased by 95.78% year-on-year. According to Trendforce Consulting Research, the decline in power batteries in the Chinese market began to converge in June, and demand recovered steadily, gradually recovering to the level of the fourth quarter of 2022. It is expected that the subsequent lithium battery materials as the upstream of power batteries may rise in volume and price, and continue to pay attention to related investment opportunities.

3. Theme and Industry Trend Changes: Focus on structured investment opportunities under the 2023 World Artificial Intelligence Conference

The 2023 World Artificial Intelligence Conference was held from July 6 to 8 at the Shanghai World Expo Center and World Expo Exhibition and Convention Center. With the theme of "Intelligent Connected World, Generating the Future", the overall architecture of the event is set according to "1+1+2+10+N", focusing on ten cutting-edge trends such as large models, smart chips, scientific intelligence, robots, brain-like intelligence, metaverse, autonomous driving, data forum, rule of law and security, and blockchain. More than 30 large models such as Baidu Wenxin, Ali Tongyi, Huawei Pangu, iFLYTEK Xinghuo, and SenseTime Ririxin were all presented, and more than 20 robots such as Cloudminds bipedal robot equipped with large model dialogue ability, Tesla humanoid robot Optimus, etc. were unveiled. In addition, the conference also conducted in-depth discussions on trends such as data element circulation, high-quality corpus data technology and rules.

(1) Robot: linear joint with roller screw as core and rotating joint with precision reducer as core

The core of the robot hardware lies in rotary and linear joints. According to the sharing by Liu Yong, chairman of Zhongbing Intelligence, at the World Artificial Intelligence Conference, there is still a gap between domestic robots and Truss robots, mainly because the accuracy of parts is not enough, and the reducer ability is not strong, resulting in relatively incoherent movements. In terms of splitting, mobility is mainly affected by rotary joints and linear joints.

Linear Joints focuses on roller screw investment opportunities. At present, with the advantage of high precision, roller screws are gradually replacing traditional trapezoidal lead screws. In the future, according to the options provided by Tesla's official website, the humanoid robot Optimus may use planetary roller screws as its transmission device. Although the planetary roller screw has a complex structure and is difficult to process, it has the advantages of strong bearing capacity, high transmission efficiency and long service life, and is the core component of future humanoid robots. At present, the development of domestic roller screws is in the early stage, limited by the accuracy of high-end machine tools, the degree of localization is low, and it is expected that the future development space is huge.

Rotating joints focus on precision reducer investment opportunities. Precision reducer is known as the "heart of industrial robots" and is the parts with the highest barriers in robot production, mainly including RV reducer, harmonic reducer and planetary reducer. At present, from the perspective of the competitive landscape, the domestic substitution rate of the reducer market is relatively low, and most of the share is still occupied by Japanese manufacturers. RV reducer giant Japan's Nabotesco accounts for about 60% of the global reducer market share, and harmonic reducer leader Hamonaco accounts for about 15% of the global reducer market share. From the perspective of domestic substitution, according to the survey of high-tech robots, the current harmonic reducer has achieved technological breakthroughs and can achieve import substitution; Due to performance problems such as transmission accuracy and torsional stiffness, the RV reducer still relies on imports.

(2) Autonomous driving: High-level autonomous driving continues to advance, following the progress of ADAS and LiDAR

Musk predicts to achieve fully autonomous driving by the end of this year. At the 2023 World Artificial Intelligence Conference, Musk said, "At present, (Tesla) is very close to achieving fully autonomous driving without human intervention,...... Full autonomous driving or level 4-5 full autonomy is expected later this year. Although this prediction cannot be 100% guaranteed, it is closer to reality than ever before. In addition, Musk revealed, "The current Tesla automatic driving is very close to the state of full automatic driving without human intervention." [Tesla] has been tested on U.S. roads and will now rarely require human intervention. ”

Domestic autonomous driving is also developing rapidly. On the commercial operation side, at the intelligent driving forum of the World Artificial Intelligence Conference, 15 vehicles of three enterprises, including Baidu Zhixing, AutoX and Pony.ai, obtained the first batch of road test licenses for driverless intelligent networked vehicles issued by Pudong New Area, and with the implementation of the "Pudong New Area Provisions on the Innovative Application of Driverless Intelligent Networked Vehicles" and relevant implementation rules, Shanghai Pudong New Area autonomous vehicles have truly ushered in the stage of unmanned; On June 17, Baidu's autonomous driving travel service platform Luobao was officially awarded a license in Pingshan District, carrying out L4 unmanned driving commercial charging operation in Pingshan District, Shenzhen, indicating that the commercialization process of China's high-level autonomous driving technology has entered the "fast lane". On the software side, on June 17, the ideal held the home technology day press conference, at which the intelligent cockpit 3M multi-modal interaction system + MindGPT cognitive large model was released, which helps to improve the vehicle's perception of the road and road objects, the prediction of the surrounding driving environment, and the decision-making power of the direction and speed of the vehicle situation, providing software support for automatic driving.

ADAS and lidar are expected to continue to benefit, and it is recommended to pay attention to related investment opportunities. In terms of ADAS, the market scale continues to expand, Yiou think tank data shows that in 2021, in the short and medium term, the market size of ADAS/autonomous driving software applications will account for 20% of the domestic automotive software market, ranking second, and it is expected that by 2025, ADAS/autonomous driving software will account for a significant proportion of cockpit software, reaching 59%, becoming the main driving force for industry growth; In terms of lidar, China is expected to become the fastest growing market, and it is expected that by 2025, the mainland lidar market size will reach 4.31 billion US dollars.

(3) Large models: Pay attention to the domestic large models represented by Huawei's Pangu model and their scene applications

Huawei released the Pangu 3.0 model. At the Huawei Developer Conference held on the afternoon of July 7, Zhang Pingan, Executive Director of Huawei, announced that HUAWEI CLOUD Pangu Model 3.0 was officially released, and Pangu Model 3.0 is divided into three layers: L0 basic model, L1 industry model, and L2 scenario model. According to reports, this is a large model that is completely oriented to the industry. Among them, Pangu L0 layer basic large model, including hundreds of billions of parameters natural language large model, Pangu multimodal large model, visual large model, prediction large model, scientific computing large model, etc., L1 layer is the industry large model, including government affairs, finance, manufacturing, mining, etc. The L2 layer is the scene model.

Positioning industry development, Huawei Pangu large models or vertical applications have more advantages. On July 6, Nature published the research results of the R&D team of HUAWEI CLOUD Pangu model, "Accurate medium-range global weather forecasting with 3D neural networks". Different from the traditional 2D neural network model, based on the Transformer model, this paper creatively proposes a three-dimensional neural network adapted to the earth coordinate system, and processes the uneven 3D meteorological data, and the results have higher accuracy than ECMWF-HRES. In May this year, the trend of Typhoon "Mawa" received widespread attention; According to the Central Meteorological Administration, HUAWEI's cloud Pangu model performed well in the path prediction of Mawa, predicting its turning path five days in advance. It is reported that this is the first paper published by Chinese technology companies as the only signing unit in recent years. The reviewer of Nature spoke highly of the results: "HUAWEI CLOUD Pangu Weather Model allows people to re-examine the future of weather forecasting models, and the opening of models will promote the development of this field. In addition to meteorology, Huawei's large model also has more advantages in mining, government affairs, etc., "not poetry, only doing things".

(4) Data elements: follow up the progress of high-quality data supply, data elementalization and data labeling

High-quality data: Data quality is the "key to success" for the model

The underlying data quality determines the output quality and application performance of large models. GIGO tells us that garbage in garbage out. The knowledge, inclusive culture, implicit thinking and hidden consciousness covered by the underlying training data will determine the capabilities and styles of large models and applications; In other words, corpus data with profanity, fallacies, extreme thoughts, and moral decay will certainly not be able to train a moral, literate, and energetic model.

At present, the stock of high-quality data is seriously insufficient. Studies have shown that stocks of high-quality linguistic data will be depleted soon. In October 2022, a team composed of MIT artificial intelligence team, Aberdeen and other university scholars based on natural language processing and computer vision methods, using machine learning models, through classification and calculation of different corpus, found that the inventory of high-quality language data will be quickly exhausted, possibly before 2026 at the earliest; Low-quality linguistic and image data take a long time, and low-quality languages may be consumed between 2030 and 2050; The stock of images may be exhausted in 2030-2060.

Data elements: solve the problems of right confirmation and revenue distribution, and effectively solve the contradiction between open data circulation and privacy security

As an "intermediate state" of data, data elements realize the form transformation of "data resources-data elements-data products". Data elements refer to data sets formed by several related fields according to market needs and application scenario needs, or data characteristics formed by modeling the associated fields of data after data desensitization processing. Specifically, 1) the original data is desensitized to generate sample data, and the sample data is used for debugging, and some algorithms need to load full data to optimize the component model; 2) Subsequently, the data element results are generated by loading the full data based on the trained component model; 3) Data elements form data products and participate in transactions.

Data Componentization helps solve the problems of compliance confirmation and revenue distribution. In terms of compliance and confirmation of rights, data elementalization follows the property rights operation mechanism of data resource ownership, data use rights, and data product operation rights; In the three data forms of data resources, data elements and data products, the aforementioned three types of property rights are confirmed and authorized. In terms of revenue distribution, in the three stages of data resources, data elements and data products, the data scale, data element value and product demand are respectively used as the basis, and the cost method, income method and market method are adopted as the pricing mechanism in turn.

Data elementalization effectively solves the contradiction between open data circulation and privacy security. The morphological change of "data resource-data element-data product" can be regarded as two irreversible chemical reaction processes, firstly, the data resource as a "reactant" generates the data element "intermediate state" under the element model processing, and secondly, the "intermediate state" data element forms the data product "final product", enters the circulation of data elements, and participates in transactions; However, the reverse reaction is not feasible, and the raw data cannot be obtained from the data product, so the data element can block the risk of critical data leakage from the inside out, as well as the risk of data tampering and data abuse from the outside in.

Data annotation: It is closely related to the high quality and security of the model

Data annotation in the AI era ensures the high quality and safety of models. Data annotation is the process of processing unstructured, unprocessed voice, picture, video and other data, so as to make it more convenient for machines to identify information; It can be said that the high quality of the data depends on the taste of the data itself on the one hand, and whether the data itself can be better absorbed by the large model on the other hand. At present, common data labeling is mainly based on image annotation, 3D point cloud annotation and other technologies. The emergence of multimodal large models puts forward higher requirements for data quality, and data annotation not only needs to reasonably label the data semantics, but also aligns different data sets of the same meaning, and adjusts parameters to make the quality of different modal data comparable to avoid the short-board effect. In addition, data labeling is closely related to data security, and it is expected that data labeling will be used to test data security factor in the future.

4. Themes and important policies

Important policies worth paying attention to last week are: (1) Zhejiang issued the Implementation Opinions on Promoting the High-quality Development of the Platform Economy, and the current focus of platform economy policies has shifted to "supporting healthy development"; (2) Beijing issued the "Implementation Opinions on Better Playing the Role of Data Elements to Further Accelerate the Development of the Digital Economy", which is the Beijing version of the "20 Data Articles", and data elements are accelerating; (3) Shanghai issued the "Several Measures to Promote the Innovation and Development of Artificial Intelligence Large Model in Shanghai", which focuses on four major directions: innovation ability, innovation elements, innovative application, and innovation environment, and builds three plans and five major platforms; (4) The Ministry of Industry and Information Technology and other relevant departments have revised the "Measures for the Parallel Management of Average Fuel Consumption of Passenger Vehicle Enterprises and New Energy Vehicle Credits", which puts forward higher requirements for the innovation and development of new energy vehicles; (5) The National Health Insurance Administration issued a draft of the renewal rules for comments, which is expected to stabilize corporate confidence and help normalize the renewal of contracts in national negotiations.

(1) Platform economy: strengthen governance and shift to promote development

Zhejiang issued the Implementation Opinions on Promoting the High-quality Development of the Platform Economy, and the platform economy policy focuses on "supporting healthy development"

On July 5, Zhejiang issued the Implementation Opinions on Promoting the High-quality Development of the Platform Economy (hereinafter referred to as the "Opinions"), proposing that "by 2027, a more dynamic innovation system, a diversified and integrated ecosystem, an accurate and efficient service system, a fair and transparent rule system and an efficient and coordinated regulatory system will be built in an all-round way. Give full play to the role of expanding the consumer market and other aspects."

Since 2022, the platform economy policy has undergone a significant transformation, from strengthening governance to supporting healthy development. In December 2020, the Central Economic Work Conference mentioned "strengthening anti-monopoly and preventing the disorderly expansion of capital", and pointed out that "it is necessary to improve the legal norms on the identification of monopoly of platform enterprises, the management of data collection and use, and the protection of consumer rights and interests." The April 2021 Politburo meeting also once again proposed "strengthening and improving the supervision of the platform economy". Since 2022, the expression of the platform economy has focused more on promoting development. The Politburo meeting in April and July 2022 and the Central Economic Work Conference in December 2022 mentioned "to promote the healthy development of the platform economy", "to promote the healthy and sustainable development of the platform economy" and "to support platform enterprises to play a role in leading development, creating jobs and international competition", and the April 2023 Politburo meeting proposed "to promote the standardized and healthy development of platform enterprises and encourage leading platform enterprises to explore and innovate".

The Opinions cover five major aspects: innovation system, ecosystem, service system, rule system and regulatory system, with a total of 25 measures, and are the first implementation opinions in China to promote the high-quality development of the platform economy. Specifically, in the construction of an innovation system, the Opinions propose to promote the construction of computing power infrastructure, optimize the layout of artificial intelligence computing power platforms, pilot the marketization of data elements in industrial manufacturing, trade logistics, social governance and other fields, and accelerate the construction of new advantages in the future industry of the metaverse. In the ecosystem, it is proposed to broaden the development space of the digital cultural industry and promote the transformation and upgrading of intelligent transportation; In the service system, it is proposed to support and guide qualified platform enterprises to list at home and abroad; In the rule system, it is proposed to establish and improve inclusive and prudent supervision rules and implement a "traffic light" system for capital investment.

(2) Digital economy: The Beijing version of the "20 Data Articles" was released, and the data elements were accelerated

The Beijing version of the "20 Data Articles" was released

On July 5, 2023, Beijing issued the Implementation Opinions on Better Leveraging the Role of Data Elements to Further Accelerate the Development of the Digital Economy (hereinafter referred to as the Implementation Opinions), proposing to "strive to achieve a market size of 200 billion yuan in the city's data element market by 2030, basically complete the pilot work of the national data infrastructure system, and form a data service industry agglomeration area".

The "Implementation Opinions" is a further implementation of the "20 Data Articles" by Beijing Municipality, and also involves more than 20 tasks, which can be regarded as the Beijing version of the "20 Data Articles". Specifically, the Implementation Opinions go a step further in the construction of the data infrastructure system, and propose to first focus on "taking the lead in implementing the data property rights and revenue distribution system", that is, exploring the operation mechanism of data rights confirmation, and taking the lead in piloting. In addition, for the first time, the Implementation Opinions link data elements to AI-generated content

In addition, according to our tracking findings, a number of policies related to data elements were introduced in June, and the implementation of data elements accelerated. From the central level, since June, the General Office of the State Council and other departments have issued two policies related to data elements to accelerate the implementation of data elements. From the perspective of local policies, according to incomplete statistics, a total of 10 provinces and cities have issued policies related to data elements since June. From the perspective of policy content, the current data element policy presents the characteristics of government-led promotion, multi-industry coverage, and accelerated marketization, mainly focusing on: 1) digital government construction and government data sharing; 2) Public data sharing such as medical insurance and industrial enterprises; 3) Digital city construction; 4) Construction of data element market and data trading platform; 5) Data element security management, among which accelerating the construction of the data element market is mentioned the most often, and the progress of data element marketization is expected to accelerate.

(3) Artificial intelligence: Shanghai issued "Several Measures to Promote the Innovation and Development of Artificial Intelligence Large Model in Shanghai" to create an AI "model" capital

Shanghai released "Several Measures to Promote the Innovation and Development of Artificial Intelligence Large Model in Shanghai"

At the closing ceremony of the World Artificial Intelligence Conference (WAIC) held on the afternoon of July 8, the "Several Measures to Promote the Innovation and Development of Artificial Intelligence Large Models in Shanghai" was announced, and the "Model" initiative was released, three batches of artificial intelligence projects were signed, and the Shanghai artificial intelligence open source ecological industry cluster was established. Zhang Ying, deputy director of the Shanghai Municipal Economic and Information Commission, introduced that the measures focus on four major directions: innovation ability, innovation elements, innovative application and innovation environment, focusing on building three plans and five major platforms.

Specifically, the three plans to solve the development bottleneck will be carried out from three aspects:

1) Large Model Innovation Support Program: Focus on supporting Shanghai innovation subjects to develop large models with international competitiveness, implement special awards, and accelerate model iteration.

2) Intelligent computing power acceleration plan: strengthen the construction of intelligent computing power of large models and establish green channels.

3) Demonstration application promotion plan: strengthen the in-depth application and benchmarking scenario of large models in vertical fields such as intelligent manufacturing, education and teaching, science and technology finance, design creativity, and scientific intelligence.

Shanghai: The first batch of science and technology off-campus training catalogs include seven categories of computer programming, robotics, model making and other projects

On July 7, the Shanghai Municipal Science and Technology Commission issued the "Catalogue of the First Batch of Science and Technology Off-campus Training Projects in Shanghai", which will be implemented from August 1, 2023. The course content mentioned in the "Catalog" includes: computer programming: including scratch, python, C/C++, wedo, JavaScript, Lingo and other courses; Robotics: including VEX robots, science and technology robots and other courses; Model making: including aircraft model, sea model, car model and other courses; Drone; Radio: including radio direction finding, radio communication and other courses; Makers; Brick building.

Shanghai has successively issued the "Shanghai Education Informatization 2.0 Action Plan", "Shanghai Education Digital Transformation "14th Five-Year Plan", "Shanghai Education Digital Transformation Implementation Plan (2021-2023)" and other documents to comprehensively promote the digital transformation of education. The issuance of the "Catalogue of the First Batch of Science and Technology Off-campus Training Projects in Shanghai" deepened the combination of artificial intelligence and education.

(4) New energy vehicles: The Ministry of Industry and Information Technology adjusted the credit calculation method for new energy vehicles, and reduced the average score of new energy passenger vehicles by about 40%.

The Ministry of Industry and Information Technology and other relevant departments have revised the "Measures for the Parallel Management of Average Fuel Consumption of Passenger Vehicle Enterprises and New Energy Vehicle Credits"

On July 6, in order to meet the needs of the development and technological progress of the mainland's energy-saving and new energy vehicle industry, the Ministry of Industry and Information Technology and other relevant departments decided to revise the "Measures for the Parallel Management of Average Fuel Consumption and New Energy Vehicle Credits of Passenger Vehicle Enterprises", which will come into force on August 1, 2023.

The revised "double credit" policy reduces the average score value of new energy passenger vehicles by about 40%, and adjusts the point calculation method and upper limit of points accordingly. A new chapter on "New Energy Vehicle Integral Pool Management" was added to establish a credit pool management system; When the supply-demand ratio of positive and negative points of new energy vehicles exceeds 2 times in the year, the credit pool storage is activated, allowing enterprises to store positive points of new energy vehicles to the credit pool on a voluntary basis, and the storage period of this part of the points is 5 years. After the revision, the "double credit" policy has further tightened the acquisition of new energy credits by car companies, put forward higher requirements for the innovation and development of new energy vehicles, and is conducive to the higher quality development of new energy vehicles.

(5) Medicine: The National Health Insurance Administration issued a draft of renewal rules for comments to help normalize the renewal of contracts in the country

The National Health Insurance Administration issued the Rules for Negotiating Drug Renewal (2023 Draft for Comments)

On July 4, the National Health Insurance Administration (CNA) issued the Rules for Negotiating Drug Renewal (2023 Draft for Comments) and the Rules for Non-Exclusive Drug Bidding (Draft for Comments). Compared with the 2022 version, the biggest change in the 2023 version is that the simple renewal part slows down the decline in drug prices, giving the market clearer expectations, which is conducive to the normalization of national contract renewal. Specifically, (1) "drugs that have been negotiated into the catalogue and have been included in the catalogue for more than 8 years in the 'negotiated drug part of the agreement period' of the catalogue are included in the routine catalogue management," which will help stabilize corporate expectations and reduce the pressure of price reduction; (2) "For varieties that have been negotiated for more than 4 years during the agreement period of the "continuous inclusion in the catalogue", the payment standard will be halved on the basis of the aforementioned calculated value," which will enhance the enthusiasm and motivation of enterprises to access medical insurance; (3) "If non-exclusive drugs are included in the medical insurance catalogue through bidding, the lowest of the quotations of each enterprise shall be taken as the payment standard for the generic drug." If the company's quotation is less than 70% of the willingness to pay by medical insurance, 70% of the willingness to pay by medical insurance is used as the payment standard for the drug, "It will be conducive to optimizing the quotation strategy of the enterprises participating in the bidding and preventing malicious competition."

5. Investment advice

In our July strategy monthly report "Economic Stabilization, Upward Offensive Continues-A-share July 23 Views and Allocation Suggestions", we proposed that we need to focus on five major tracks with marginal improvement: integrated circuits (storage, packaging and testing, materials), digital economy (data elements), new energy vehicles (vehicles, batteries, autonomous driving), AI+ (applications, algorithms, computing power), and photovoltaics (N-type modules). The latest marginal improvements in the five recent tracks are as follows:

From a medium- to long-term perspective, we recommend paying attention to the rise of the whole industry chain of carbon neutrality (photovoltaic, wind power, energy storage, hydrogen energy, etc.) under the global attention to carbon neutrality, the increase in the penetration rate of electric smart vehicles under the general trend of electrification and intelligence, and industrial trends such as digital economy, artificial intelligence, metaverse, and autonomous controllability.

02 Global Watch: Global Market Tracker

1. The performance of the global stock market industry

Global stock markets rose and fell more last week, with utilities and optional consumer sectors performing relatively well, and healthcare and information technology performing relatively poorly.

Specifically, in terms of U.S. stocks, the real estate and utilities sectors performed relatively well, while the healthcare and materials sectors performed relatively poorly, and the overall poor performance of U.S. stocks was mainly due to the hawkish comments in the FOMC minutes indicating that the Fed will continue to raise interest rates this year. European stocks fell, the energy and telecommunications services sectors performed relatively well, the industrial and optional consumer sectors performed relatively poorly, and the Eurozone PPI turned negative for the first time since May 2020, and deflationary pressures gradually became prominent. The UK stock market fell generally, with the materials and energy sectors performing relatively well, and the healthcare and industrial sectors performing relatively poorly. In the Japanese stock market, the financial and energy sectors performed relatively well, while the healthcare and materials sectors performed poorly. In terms of Hong Kong stocks, the optional consumption and energy sectors performed relatively well, while the utilities and financial sectors performed relatively poorly. In terms of A-shares, the energy and utilities sectors performed relatively well, while the information technology and healthcare sectors performed relatively poorly.

2. Brief analysis of global strong stocks and variable stocks

Among the $20 billion companies with a market value that rose sharply in the past week, most of them are companies in the energy and optional consumer industries; Among the $20 billion companies with market value that led the decline in the past week, most of them were companies in the information technology and health care sectors.

A brief analysis of the change unit

【Tesla】Tesla produced 479,700 vehicles in the second quarter, and delivered 466,100 units, setting a new quarterly delivery record for the company. MODEL 3/Y WAS PRODUCED 460,200 UNITS WITH 446,900 UNITS DELIVERED, WHILE MODEL S AND MODEL X WERE 19,500 UNITS AND 19,200 UNITS.

AstraZeneca's findings on the new cancer drug datopotamab deruxtecan (Dato-DXd) have raised concerns that the drug may not work as well as expected. AstraZeneca said July 3 that the drug helped patients with the most common lung cancers live longer than standard chemotherapy and did not show deterioration. However, AstraZeneca has not yet given a clearer picture of the trial's progression-free survival and overall longevity.

03 Important information delivery

1. Domestic industrial policy and news

(1) Digital economy: MIIT: Accelerate the innovation and development of strategic emerging industries such as big data, artificial intelligence, and intelligent networked vehicles (☆☆☆)

On July 4, the 2023 Global Digital Economy Conference opened in Beijing. Vice Minister of Industry and Information Technology Wang Jiangping attended the opening ceremony and delivered a speech. Wang Jiangping said that the Ministry of Industry and Information Technology will accelerate the deep integration of the digital economy and the real economy, and coordinate to promote digital industrialization and industrial digitalization, so that the digital economy can better empower the entity, serve the society and benefit the people. Accelerate the construction of new infrastructure, promote the systematic development and large-scale deployment of digital infrastructure, and accelerate the construction of high-speed ubiquitous, space-ground integration, cloud-network integration, intelligent and agile, green and low-carbon, safe and controllable intelligent comprehensive digital information infrastructure. Promote the digital transformation of industries, vigorously promote the integrated application of new generation information and communication technologies such as 5G and gigabit optical networks in vertical industries, information consumption, social and people's livelihood, and form a demonstration benchmark for innovative applications in key fields. Enhance the momentum of digital development, accelerate the innovation and development of strategic emerging industries such as big data, artificial intelligence, and intelligent networked vehicles, vigorously develop advanced manufacturing, promote the digital economy to become stronger, better and bigger, and cultivate new economic engines.

(2) Data elements: Beijing: Promote the research and development of ultra-large-scale intelligent models such as natural language, multi-modality, and cognition, and strive to reach 200 billion yuan in the data element market by 2030 (☆☆☆)

On July 5, Beijing issued the Implementation Opinions on Better Leveraging the Role of Data Elements to Further Accelerate the Development of the Digital Economy, involving 20 specific tasks. The implementation opinions put forward the formation of a number of data systems, policies and standards that are carried out in advance and piloted. Promote the establishment of a multi-level data trading market that efficiently matches supply and demand, fully tap the value of data assets, and build a hub highland for data element allocation. Promote the open development of the entire industry chain of the digital economy and international exchanges and cooperation, form a number of data-empowered innovative application scenarios, and cultivate a group of leading enterprises based on data elements. By 2030, the scale of Beijing's data element market will reach 200 billion yuan, basically complete the pilot work of the national data infrastructure system, and form a data service industry agglomeration area. Support the Beijing Economic and Technological Development Zone and other countries to carry out data infrastructure systems and pilot projects, and build policy highlands, trusted spaces, and data factories. Promote national pilot projects of data intellectual property rights, exploring the establishment of systems, registration practices, protection of rights and interests, and transaction use of data intellectual property rights. Establish social data asset registration centers, establish data asset assessment service stations, and first explore and carry out data asset entry into tables. Build a data element innovation institute to support data-driven scientific research. Improve the artificial intelligence data annotation library, explore the creation of a data training base, and promote the research and development of ultra-large-scale intelligent models such as natural language, multi-modality, and cognition.

(3) Semiconductors: Ministry of Commerce, General Administration of Customs: export control of gallium and germanium-related items (☆☆☆)

(4) Satellite communication: Mainland successfully launched satellite Internet technology test satellite (☆☆☆)

Jiuquan, 9 Jul (Xinhua) -- At 19:00 on 9 July, the mainland successfully launched the satellite Internet technology test satellite using the Long March-2C carrier rocket at the Jiuquan Satellite Launch Center. The mission was the 478th flight of the Long March series of launch vehicles.

(5) Large model: HUAWEI CLOUD Pangu Model 3.0 officially released (☆☆☆)

At the Huawei Developer Conference 2023 (Cloud) held on the afternoon of July 7, Zhang Pingan, Executive Director of Huawei and CEO of HUAWEI CLOUD, announced the official release of HUAWEI CLOUD Pangu Model 3.0. Zhang Pingan said that Pangu Big Model 3.0 is divided into a three-layer architecture: L0 basic large model, L1 industry large model, and L2 scene model, which will reshape thousands of industries. According to reports, this is a large model that is completely oriented to the industry. Among them, Pangu L0 layer basic large model, including hundreds of billions of parameters natural language large model, Pangu multimodal large model, visual large model, prediction large model, scientific computing large model, etc., L1 layer is for the industry large model, including government affairs, finance, manufacturing, mining, etc. The L2 layer is the scene model.

(6) Large model: The research results of HUAWEI CLOUD Pangu Meteorological Big Model were published in the official journal of Nature (☆☆)

On July 6, Nature published the research results of the R&D team of HUAWEI CLOUD Pangu model, "Accurate medium-range global weather forecasting with 3D neural networks". The data shows that this is the first published paper of Nature by a Chinese technology company as the only signatory in recent years. The reviewer of Nature spoke highly of the results: "HUAWEI CLOUD Pangu Weather Model allows people to re-examine the future of weather forecasting models, and the opening of models will promote the development of this field. ”

(7) Autonomous driving: Shanghai Pudong New Area intelligent networked vehicles without drivers officially launched on the road, Baidu Zhixing and other three companies won the first batch of road test cards (☆☆)

On July 8, with the implementation of the "Pudong New Area Regulations on the Innovative Application of Intelligent Networked Vehicles without Drivers" and relevant implementation rules, self-driving cars in Shanghai Pudong New Area truly ushered in the stage of unmanned. At the 2023 World Artificial Intelligence Conference - Intelligent Driving Forum held at Zhangjiang Science Hall on July 8, 15 vehicles from three companies, including Baidu Zhixing, AutoX and Pony.ai, obtained the first batch of road test licenses for intelligent networked vehicles without drivers issued by Pudong New Area.

(8) New energy vehicles: The Ministry of Industry and Information Technology adjusted the calculation method of new energy vehicle points, and reduced the average score of new energy passenger vehicles by about 40% (☆☆)

On July 6, the Ministry of Industry and Information Technology and four other departments revised the "Measures for the Parallel Management of Passenger Car Enterprise Average Fuel Consumption and New Energy Vehicle Credits", adjusting the calculation method of new energy vehicle credits, which came into effect on August 1, 2023. The Measures reduce the average score of new energy passenger vehicles by about 40%, and adjust the point calculation method and upper limit of points accordingly. A new chapter on "New Energy Vehicle Points Pool Management" has been added to establish a credit pool management system. When the supply-demand ratio of positive and negative points of new energy vehicles exceeds 2 times in the year, the credit pool storage is activated, allowing enterprises to store positive points of new energy vehicles to the credit pool on a voluntary basis, and the storage period of this part of the points is 5 years. When the supply-demand ratio of positive and negative points of new energy vehicles in the year does not reach 1.5 times, the points in the point pool are released, allowing enterprises to withdraw the stored positive points of new energy vehicles, and the positive points of unused new energy vehicles in the current year will be returned to the point pool.

(9) Platform economy: Zhejiang issued the Implementation Opinions on Promoting the High-quality Development of Platform Economy (☆☆)

On July 5, Zhejiang issued the "Implementation Opinions on Promoting the High-quality Development of the Platform Economy": 1) Encourage platform enterprises to form innovation consortiums with upstream and downstream enterprises in the industrial chain and colleges and universities, organize scientific research around the underlying technology of the platform economy and "stuck neck" technology, and achieve a number of major landmark achievements in the fields of big data, cloud computing, blockchain, operating systems, and processors. Promote the construction of computing power infrastructure, optimize the layout of artificial intelligence computing power platforms, strengthen algorithm innovation and application, and build an algorithm transformation and application ecology. 2) Promote the construction of an open and shared data platform. Platform enterprises are encouraged to build open and shared cloud platforms to provide small and medium-sized enterprises with services such as on-demand use, dynamic expansion, high-quality and low-cost data storage, software development, and transaction payment. Promote the formulation of standards for system migration and interconnection between cloud platforms to accelerate business and data interconnection. Platform enterprises are encouraged to provide low-code-driven digital platform tools, and promote operators within the platform to improve network operation efficiency through technology empowerment. Support platform enterprises to promote the construction of software open source communities and ecological alliances, and drive a number of small and medium-sized software enterprises to integrate into open source. Continue to advance the open sharing of public data, and establish standardized, orderly, secure, and efficient mechanisms for the development and utilization of public data.

(10) Artificial intelligence: Shanghai issued "Several Measures to Promote the Innovation and Development of Artificial Intelligence Large Model in Shanghai" to create an AI "model" capital (☆☆)

At the closing ceremony of the World Artificial Intelligence Conference (WAIC) held on the afternoon of July 8, the "Several Measures to Promote the Innovation and Development of Artificial Intelligence Large Models in Shanghai" was announced, and the "Model" initiative was released, three batches of artificial intelligence projects were signed, and the Shanghai artificial intelligence open source ecological industry cluster was established. Zhang Ying, deputy director of the Shanghai Municipal Economic and Information Commission, introduced that the measures focus on four major directions: innovation ability, innovation elements, innovative application and innovation environment, focusing on building three plans and five major platforms. The three plans to solve the bottleneck of development, one is the large model innovation support plan, which focuses on supporting Shanghai innovation subjects to develop large models with international competitiveness, implement special awards, and accelerate model iteration; The second is the intelligent computing power acceleration plan, strengthening the construction of intelligent computing power of large models and establishing green channels; The third is the demonstration application promotion plan, strengthening the in-depth application and benchmarking scene creation of large models in vertical fields such as intelligent manufacturing, education and teaching, science and technology finance, design creativity, and scientific intelligence.

(11) Photovoltaic: Silicon Industry Branch: Polysilicon prices stopped falling and stabilized this week, and N-type polysilicon rebounded (☆☆)

On July 5, the Silicon Industry Branch of the China Nonferrous Metals Industry Association released data that the domestic N-type material price range this week was 7.2-80,000 yuan / ton, and the average transaction price was 74,200 yuan / ton, a week-on-week increase of 0.68%; The price range of monocrystalline dense material was 6.00-74,000 yuan / ton, and the average transaction price was 65,700 yuan / ton, which was flat week-on-week. The overall polysilicon market stabilised this week, and N-type polysilicon rebounded.

(12) Robots: The 2023 World Robot Conference will be held on August 16, and more than 50 new products will be unveiled (☆☆)

The 2023 World Robot Congress will be held in Beijing Yizhuang from August 16 to 22. According to reports, this year's robot conference is co-sponsored by the Beijing Municipal People's Government, the Ministry of Industry and Information Technology, and the China Association for Science and Technology, with the theme of "open innovation and sharing the future", with three major sections: forums, exhibitions and competitions. Among them, the Robot Expo for the first time innovatively planned the whole museum as a "robot +" application scenario for display, covering 10 scene application sectors and core components such as robot + manufacturing, agriculture, and construction, and highlight exhibits such as bionic swift robots, four-legged robot dogs, robotic fish aquariums and more than 50 first-time new products will be unveiled.

(13) New energy vehicles: Passenger car retail sales in June were 1.896 million units, NEV retail sales were 638,000 units, up 19% y/y (☆)

On July 5, the China Passenger Vehicle Association announced the sales data of passenger cars in June, and from June 1 to 30, the passenger car market retailed 1.896 million units, down 2% year-on-year and up 9% month-on-month. Since the beginning of this year, the cumulative retail sales have reached 9.528 million units, a year-on-year increase of 3%. According to preliminary statistics from the Passenger Association of China, the retail sales of new energy vehicles from June 1 to 30 were 638,000 units, a year-on-year increase of 19% and a month-on-month increase of 10%; Since the beginning of this year, the cumulative retail sales have reached 3.059 million units, a year-on-year increase of 36%.

(14) Innovative drugs: The National Health Insurance Administration publicly solicits opinions on the "Rules for Negotiating Drug Renewal" and "Non-exclusive Drug Bidding Rules" (☆)

On July 4, the National Health Insurance Administration publicly solicited opinions on the Rules for Negotiating Drug Renewal and the Rules for Non-exclusive Drug Bidding. Among them, the Rules for Negotiating Drug Renewal Propose that drugs that meet one of the following conditions can be included in the regular catalogue management: non-exclusive drugs; Drugs negotiated in the 2019 catalogue, exclusive drugs that have not adjusted the payment standards and payment scope for two consecutive agreement cycles; Drugs that have been negotiated into the catalogue and have been included in the "Negotiated Drug Part" of the catalogue for more than 8 consecutive years. The conditions of the Non-Exclusive Drug Bidding Rules are: after expert review, it is recommended to add non-exclusive drugs to be included in the medical insurance drug list. Except for selected drugs and government-priced drugs for centralized procurement of drugs organized by the state. In addition, the "Non-exclusive Drug Bidding Rules" mentions that if a drug is included in the medical insurance catalogue through bidding, the lowest of the quotations of each enterprise shall be taken as the payment standard for the generic drug. If the company's quotation is less than 70% of the willingness to pay by medical insurance, 70% of the willingness to pay by medical insurance shall be used as the payment standard for the drug.

2. Foreign industrial policy and news

(1) Artificial intelligence: OpenAI announced that the latest large-scale language model, GPT-4, will be officially open for use through its API (☆☆☆)

On July 6, local time, OpenAI announced that its latest large-scale language model, GPT-4, will be officially opened for use through its API. Currently, all OpenAI's API paying users (those with a history of successful payments) have access to GPT-4 with 8K contextual memory. And the company plans to open up API access to new developers by the end of this month, and then adjust the limits available through hashrate availability. This means that developers around the world can use GPT-4 large-scale language models to enhance their own applications or develop entirely new generative AI applications.

(2) Solid-state batteries: Toyota: Major breakthroughs in solid-state battery technology can halve battery cost and weight (☆☆☆)

On July 4, Toyota said it had made a major breakthrough in solid-state battery technology, capable of halving the weight, volume and cost of batteries, which could give a huge boost to the advancement of electric vehicles; The production process of materials used to make solid-state batteries has been simplified, and the discovery is a major leap forward, dramatically reducing car charging time and increasing range. Keiji Kaita, president of Toyota's Carbon Neutral R&D Center, said the company has developed ways to improve battery durability and believes it can now make solid-state batteries with a range of 1,200 kilometers and a charging time of 10 minutes or less. According to the Financial Times, the company expects to start producing solid-state batteries for electric vehicles in 2027.

(3) Humanoid robots/autonomous driving: Musk: The number of robots on Earth exceeds the number of humans seems to be the development trend It is expected that full autonomous driving will be realized around the end of this year (☆☆☆)

On July 6, Tesla CEO Musk "appeared" at the 2023 World Artificial Intelligence Conference to talk about the humanoid robot Optimus, automatic driving, artificial intelligence, etc. He said that in the future, there will be more robots on Earth than there are humans. Tesla is happy to share self-driving technology with automakers. Musk predicts that with the rapid development of artificial intelligence technology, full autonomous driving will be achieved around the end of this year. "I've made many similar predictions before, and I admit that the previous predictions weren't entirely accurate, but this time, I think the predictions are relatively close."

(4) Artificial intelligence: NVIDIA's acquisition of AI startup OmniML is expected to accelerate the use of AI on edge devices (☆☆)

On July 4, NVIDIA is further enriching its AI ecosystem through acquisitions, according to recent revelations, NVIDIA secretly acquired AI start-up OmniML in February. OmniML is a company that specializes in shrinking machine learning models to edge devices, enabling machine learning tasks to be 10x faster on different edge devices with 1/10th the engineering effort. Last year, it launched a platform called Omnimizer that adapts compute-intensive AI models to low-end hardware, resulting in smaller, faster, and better suited models for the hardware it runs on.

(5) Silicon carbide: Wolfspeed, the global leader in silicon carbide, signed a 10-year silicon carbide wafer supply agreement with Renesas Electronics (☆)

On July 5, Renesas signed a 10-year silicon carbide wafer supply agreement with Wolfspeed. Under the agreement, Renesas has paid Wolfspeed a $2 billion deposit to secure the supply of SiC wafers and support Wolfspeed's U.S. capacity expansion plans.

(6) Carbon emissions: IMO proposes the latest emission reduction strategy: strive for the global shipping industry to achieve net zero emissions around 2050 (☆)

On July 8, IMO member states stressed in the 2023 Ship Greenhouse Gas Reduction Strategy adopted on July 7 that greenhouse gas emissions from international shipping should be peaked as soon as possible, and net zero greenhouse gas emissions should be achieved before or around 2050, taking into account different national circumstances.

3. Global key technology company tracking

- END -

What are the thematic investment opportunities at the World Artificial Intelligence Conference 2023? -Global Industry Trends Weekly Report (0703)

"Robots are expected to become the ultimate carrier of large models, top-level design promotes medical insurance data sharing-Global Industry Trend Weekly Report (0626)" "Pay attention to the "data lithiumization" industry trend in the AI+ era-Global Industry Trend Weekly Report (0619)"

"New energy and artificial intelligence support increases, database construction of artificial intelligence new barriers-Global Industry Trend Weekly Report (0612)"

"Huawei Builds AIoT Intelligent Ecosystem, National Standing Committee Proposes to Consolidate and Expand the Advantages of New Energy Vehicles-Global Industry Trend Weekly Report (0605)"

"C919 successfully completed its commercial first flight, NVIDIA provides customized game AI foundry - Global Industry Trend Weekly Report (0529)"

"The rise of the wave of humanoid robots, ChatGPT landed on the iOS system - Global Industry Trend Tracking Weekly Report (0522)" "Apple MR is expected to become a phenomenal terminal, ChatGPT Plugin officially opened - Global Industry Trend Weekly Report (0515)" ""China Special Valuation" leads the revaluation of the value of low-valued central enterprises, charging piles welcome the top layer of help-Global Industry Trend Weekly Report (0508)" "Condensed matter batteries have both performance and safety, diplomacy and exports dual drive the "Belt and Road" Welcoming Continuous Catalysis-Global Industry Trend Tracking Weekly Report (0424)", "The First "West-to-East Transmission" Pipeline Included in National Planning, "Artificial Sun" Achieves Major Breakthroughs-Global Industry Trend Tracking Weekly Report (0417)", "Domestic AI Large Model Ready to Launch, SAM Model Opens Machine Vision GPT Era-Global Industry Trend Weekly Report (0410)"

"OpenAI enters humanoid robots, AI regulation and autonomous controllability are imminent-Global Industry Trend Tracking Weekly Report (0403)"

"Technology giants rush to enter AI+ medical care, and the first solid hydrogen power generation in China is connected to the grid - Global Industry Trend Tracking Weekly Report (0327)"

This article is selected from the Brokerage Research Report