Back to AI News Briefing
24-Hour AI News: Policies and Models Advance in Harmony, Global Computing Power and Regulation Synchronize

24-Hour AI News: Policies and Models Advance in Harmony, Global Computing Power and Regulation Synchronize

AI News Briefing Admin 59 views

In the past 24 hours, there have been concentrated industrial policy and industry model dynamics in China, with local action plans and university/institute model iterations being promoted simultaneously; overseas, the focus has been on computing power and data center construction, enterprise-level AI ecosystem expansion, and child safety legislation, and the pace at home and abroad is generally balanced.

1. Shanghai Releases Smart Terminal Action Plan (2026-2027)

  1. Released on October 14, it proposed that the scale of the smart terminal industry will exceed 300 billion yuan by 2027.
  2. Quantitative target: The annual production capacity of AI computers, AI mobile phones, and AI new terminals will reach more than 10 million units each.
  3. The direction focuses on end-edge-cloud collaboration, multimodal human-machine collaboration and the construction of intelligent computing network carrying capacity.

2. China Agricultural University releases Shennong Model 3.0

  1. It was officially released at the World Agricultural Science and Technology Innovation Conference held in Beijing on October 13.
  2. Positioned as an iterative version for agricultural scenarios, it emphasizes multidisciplinary integration and scenario-based applications.
  3. All parties present exchanged views on the implementation of agricultural technology and models, and promoted collaboration between industry, academia and research.

3. Zidong Taichu 4.0 released, iterative upgrade of multimodal reasoning

  1. Official reports on October 14th showed that Zidong Taichu has completed its fourth major iteration to 4.0.
  2. Jointly developed by the Institute of Automation of the Chinese Academy of Sciences and the Wuhan Institute of Artificial Intelligence, it emphasizes multimodal reasoning capabilities.
  3. Cross-modal understanding and generation applications for general and multi-industry scenarios.

Baidu World 2025 is scheduled for November 13 (Beijing)

  1. It was officially announced on October 13 that the theme is "Emergence of Effects" and it will be held at the National Convention Center.
  2. Set up a "1+6" structure, focusing on AI applications, big models, AI ecology and overseas expansion.
  3. According to industry sources, the company may disclose its overseas expansion progress and full-stack application practices.

5. "AI+" Leads Collaborative Transformation of Digital Industries (Wuhan)

  1. Local media reported on October 14: Digital and intelligent collaborative transformation will become an important path for industrial upgrading.
  2. Trend highlights: Intelligent computing centers become infrastructure, and industry-level data sets and large and small models are coordinated and accelerated.

6. OpenAI and Broadcom announce 10GW self-developed AI accelerator collaboration

  1. Released on October 13, OpenAI designed its own accelerator and system, and Broadcom participated in the development and deployment.
  2. The goal is to embed cutting-edge model experience directly into hardware to improve the efficiency of large-scale intelligent systems.
  3. After the "6GW-level" cooperation in early October, the computing power ecosystem has been expanded to the "10GW-level".

7. California Signs AI Chatbot Disclosure and Child Safety Act

  1. SB-243 was signed on October 13, requiring AI to be clearly labeled in interactions that could be mistaken for humans.
  2. For scenarios involving minors and suicide risks, relevant operators must submit annual safety reports starting from 2026.
  3. This bill, together with the accompanying child online safety initiatives, promotes AI compliance and user protection.

8. Google announced a $15 billion investment in India over the next five years to build an AI hub

  1. On October 14, it was announced that a large AI center and a gigabit-class data center cluster would be built in Visakhapatnam.
  2. The plan includes submarine international gateways, energy infrastructure and fiber optic network expansion, and is in collaboration with the Adani Group.
  3. Aims to drive India’s digital transformation and regional computing power supply.

9. Salesforce will invest $15 billion in San Francisco to advance its AI ecosystem

  1. On October 13, a five-year investment plan was announced to build an AI incubation and application promotion hub.
  2. Cooperate with the Agentforce 360 platform and ecological initiatives to implement enterprise-level AI scenarios.
  3. It is expected to boost exhibitions and employment, coupled with its global investment layout.

10. Nvidia: Meta and Oracle adopt Spectrum-X Ethernet switches for AI data centers

  1. On October 13, NVIDIA announced that Meta will introduce related switches into the open source switching system platform.
  2. Oracle will use this to build a "giga-scale" AI supercomputing network.
  3. Demonstrates the increasing adoption of Ethernet technology in linearly scalable AI data centers.

Frequently Asked Questions (Q&A)

Q: How is the 10GW collaboration between OpenAI and Broadcom different from the previous 6GW collaboration?

A: The timing and partners differ: October 13th marked the development and deployment of OpenAI's self-developed accelerator with Broadcom, with a scale of 10GW; October 6th marked the 6GW collaboration with AMD. This new collaboration emphasizes the "self-developed design + system-level" approach and a higher scale.

Q: What specific numbers does the Shanghai Action Plan give for AI terminals?

A: By 2027, the scale of the smart terminal industry will exceed 300 billion yuan, with the annual production capacity of AI computers, AI mobile phones and AI new terminals reaching more than 10 million units each, and more than 3 consumer terminal brands with global influence and 2 leading companies will be cultivated.

Q: What are the core obligations of AI chatbots under California’s new bill?

A: In interactions that may be mistaken for humans, the "AI identity" must be clearly marked; operators involving minors and mental health scenarios must submit annual disposal and response reports starting in 2026 and disclose the key points to the public.

Q: What infrastructure is Google building for its $15 billion AI hub in India?

A: Large data centers, international submarine gateways, energy and fiber optic network expansion, etc., and will be promoted in collaboration with the Adani Group over the next five years.

October 13-14 AI News Shanghai Smart Terminal Action Plan 2027 Smart Terminals: 300 billion yuan Annual production capacity of AI computers: tens of millions units Annual production capacity of AI mobile phones: tens of millions units Device-edge-cloud collaborative layout Multimodal human-machine collaboration Intelligent computing network construction Shennong Model 3.0 released Multidisciplinary integration of AI in agriculture Scenario-based application of agricultural models Zidong Taichu 4.0 Multimodal Reasoning Jointly developed by the Institute of Automation, Chinese Academy of Sciences Cross-modal understanding and generation Baidu World 2025 release date set Effect emergence AIinAction 1+6 Forum Structure AI applications and overseas development progress Wuhan Artificial Intelligence + Transformation Intelligent Computing Center Infrastructure Industry-level data sets implemented OpenAI and Broadcom 10GW Collaboration Self-developed AI accelerator system 10GW computing power ecosystem expansion California AI Chatbot Disclosure Law Child Online Safety Legislation 2026 Annual Security Report Requirements Google's $15 billion AI hub in India Visakhapatnam Data Center Construction of submarine international gateway Partnering with Adani Group Salesforce invests $15 billion in San Francisco Agentforce360 Ecosystem Expansion Enterprise-level AI incubation hub NVIDIA Spectrum-X switch Meta adopts open source exchange platform Oracle's giga-scale supercomputing network The trend of multipolarization of global computing power Child Safety and AI Compliance Dual drive from policy and industry Quantified target for AI terminal production capacity End-side AI productization Accelerated iteration of industry models Agricultural AI Digital Governance Enterprise-level AI ecosystem expansion Data Center and Energy Layout Intelligent Agents and Application Popularization Localized computing power and network Synergy between capital and computing power Dynamic balance of AI at home and abroad

Recommended Tools

More