24-hour AI news: policy refinement and scenario implementation go hand in hand, and overseas platforms are betting on model synergy

24-hour AI news at a glance: the implementation of domestic "artificial intelligence +" policies is refined, and the exhibition and industrial applications are intensively updated; In

past 24 hours, domestic "artificial intelligence +" supporting measures have been promoted at the same frequency as industrial activities, and local and ministries have clearly avoided disorderly competition. A number of large-scale exhibitions and industry activities focused on the implementation of AI. Overseas, Meta was exposed to negotiations to access external large models and urgently strengthen the safety of teenagers, the stock prices of AI server and chip-related companies fluctuated, and India and Japan accelerated their collaborative cooperation in AI and semiconductors. In order to avoid duplication, this issue has completed the deduplication check of the content of the designated channel in the past 72 hours, and no incidents of the same topic that need to be eliminated have been found.

1. The National Development and Reform Commission refines the "artificial intelligence +" action: emphasizing overall planning and avoiding disorderly competition

1. On August 29, the authoritative interpretation proposed that it will work with relevant parties to formulate specific implementation plans for "artificial intelligence +" by industry, focusing on six major actions of science and technology, industry, consumption, people's livelihood, governance, and global cooperation.
2. The caliber is clear to "avoid disorderly competition and rush up", emphasizing the joint efforts of the national chess game and policies.
3. The stage goal proposes that the next 1-2 years will be the key window period for landing, and the penetration rate and industrial scale will be jumped in stages by 2027 and 2030 respectively.

2. The first artificial intelligence exhibition of the China-Arab States Expo: Yinchuan builds a new window for China-Arab AI cooperation

1. The 7th China-Arab States Expo was held from August 28 to August 31, and an artificial intelligence exhibition area was set up for the first time.
2. The exhibition area focuses on new technologies and applications, serving industry docking and international cooperation.
3. Connect with the "artificial intelligence +" action and expand AI scenarios and ecological cooperation with the "Belt and Road" countries.

3. The 2025 China International Big Data Industry Expo opens in Guiyang: AI applications are unveiled

1. It will open on August 30 for 4 days, with 6 themed pavilions set up to attract more than 300 enterprises to participate in the exhibition.
2. The exhibition covers multi-scenario applications such as large models, embodied intelligence, unmanned driving and AI logistics.
3. With the theme of "Digital Gathering Industry Momentum and Intelligent Development", emphasizing the integration and empowerment of data and AI.

4. The 2025 Chengdu Auto Show focuses on self-developed chips and VLA large models: car companies enter the "dividend period" of cost control 1

. On August 30, a number of new forces focused on displaying self-developed chips, on-board large models and humanoid robots.

1. The technical strategy of car companies has shifted from "new models" to "cost reduction and efficiency increase and computing power/model implementation".
2. AI R&D investment accelerates the return to scenarios and cost benefits, and promotes intelligent driving and cockpit iteration.

5. The progress of Baidu's intelligent cloud product line update has been continuously watched: Qianfan 4.0 and Baige 5.0 accelerate the landing of enterprises

1. Industry reports say that "Sight 5.0" supports "one-sentence generation of professional-grade visual AI applications", compressing weekly development to minutes.
2. The platform layer (PaaS) and the application layer (SaaS) are strengthened simultaneously to highlight standardization and replicable capabilities.
3. Echoing the "artificial intelligence +" policy, accelerate the implementation of manufacturing, quality inspection, safety production and other industries.

6. Meta was exposed to discuss model access with Google and OpenAI: a "stopgap combination" before self-developed Llama 5

1. On August 30, a number of media outlets relayed that Meta is exploring accessing Gemini and OpenAI models in its applications to enhance the experience of products such as "Meta AI".
2. Positioned as a temporary strategy before the maturity of the self-developed next-generation model Llama 5, while maintaining the open source and self-developed route.
3. Reflect the competitive strategy of large platforms on the supply side diversification and "time for space".

7. Meta adds youth safety protection: restricting AI roles and training restricted areas

1. From August 29th to 30th, Meta launched stricter youth interaction safety rules under the pressure of public opinion and regulation.
2. Temporarily restrict teenagers from accessing some AI characters, and introduce "risk avoidance" boundaries in training and recovery strategies.
3. Youth AI safety governance has become one of the core issues of platform compliance and branding in Europe and the United States.

8. AI server leader fluctuations: Super Micro is under pressure due to weakening internal control concerns

1. On August 29-30, it was reported that Super Micro's stock price fell due to concerns related to financial internal control.
2. The market's sensitivity to AI server costs, channels and competition has increased, and the overall sentiment of the sector is under pressure.
3. Show the cyclical game between "AI infrastructure prosperity, valuation digestion, and performance realization".

9. U.S. technology stocks are dragged down by AI uncertainty: adjustment of chips and related targets

1. The market review on August 30 showed that chips and some "AI concept" targets weakened, and Marvell, Nvidia, etc. were under pressure.
2. Investors re-evaluate the pace of AI spending, order sustainability and valuation tolerance.
3. Reflect that the "AI investment cycle" has entered a critical stage of switching from story to cashing.

10. India and Japan announced a roadmap for cooperation between AI and semiconductors: signed a number of agreements

1. On August 30, a number of media reported that India and Japan reached a roadmap for cooperation covering AI, semiconductors and human capital training.
2. Carry out medium and long-term cooperation around LLM training, industrial chain collaboration and security resilience.
3. Under the trend of regional industrial cooperation and "friend-shore outsourcing", the AI-chip ecology in Asia is accelerating the reconstruction.

Frequently Asked Questions (Q&A)

Q: What is the most critical landing point of the "AI+" initiative in the near future?

A: The implementation plan by industry will be implemented one after another, and the window period will be the next 1-2 years; By 2027 and 2030, the stage goals of penetration rate and industrial scale will be completed respectively, emphasizing overall planning, coordination and avoiding disorderly competition.

Q: Why did Meta consider accessing external models instead of "ALL IN self-development"?

A: In order to shorten the capability gap and product iteration cycle, Llama 5 will be completed through cooperation before it matures, while maintaining the parallel two lines of self-development and open source to reduce the risk of single-point technology and schedule.

Q: What does intensive domestic exhibitions and industry activities mean for the implementation of AI?

A: With the increase in "visible scenarios" at both ends of supply and demand, capital and enterprises will pay more attention to unit cost, ROI and delivery cycle, and promote the whole chain of collaboration of models, computing power, data and applications.

Q: Why do the AI server and chip sectors fluctuate in stages?

A: High costs, high competition and uncertain order rhythm superimposed on internal control and disclosure matters have prompted the market to repricing valuation and cashing into a "verification and differentiation period".