IRDH275B-435  • In the future, AI agents will reconstruct the human-computer interaction mode, embodied intelligence will become the next wave of subverting the manufacturing industry, intelligent computing centers will be connected, artificial intelligence computing power network coordination and acceleration of deployment, and “model as a service” will become the core of the construction of AI industry ecology.
• Composite application-oriented talents such as cross-field talents who can actively use AI large models to empower business, AI algorithm talents who can develop application scenarios, and management and decision-making talents who understand AI large models will become the main talent needs in the field of artificial intelligence.

AI agents will reconstruct the human-computer interaction mode, embodied intelligence will become the next wave of subverting the manufacturing industry, and “model as a service” will become the core of the construction of AI industry ecology.

IRDH275B-435  On January 16, at the 2024 CCID Consulting IT Trend conference, Zou Debao, executive deputy general manager of CCID Consulting Artificial Intelligence and Big Data Research Center, made the above analysis on the trend of artificial intelligence industry with the title of “2024 Artificial Intelligence Industry Trend Insight: Paving the” fast track “of digital economy innovation and development”. He believes that China’s artificial intelligence industry will make great progress in the next 10-15 years, and it is expected that by 2035, the scale of China’s artificial intelligence industry will be 1.73 trillion yuan, accounting for 30.6% of the global total.

Artificial intelligence has become a major driving force for the innovation and development of the digital economy. Zou Debao said that AI has taken the lead in releasing industrial value in digital content and media, financial information IRDH275B-435  services, digital retail, smart medical care and smart manufacturing. Next, there will be ten trends in the AI industry.

First, China’s artificial intelligence industry will make great progress in the next 10-15 years, and a number of industrial elements will lead the world. With the rapid development of the digital economy industry, the integration of China’s artificial intelligence industry and the real economy is accelerating, and it is expected that by 2035, China’s artificial intelligence industry will complete the transformation from the demonstration application exploration period to the scale IRDH275B-435  application maturity period.

Second, the “digital midpoint” will arrive 10 years earlier with the blessing of generative AI. The “digital midpoint” refers to the point in time when the digitization rate reaches 50%. Zou Debao said that it is expected that by 2035, the digitalization rate will exceed 85%, which will further promote a new way of working, improve the commercial efficiency of enterprises, comprehensively build human production and life style in the era of digital economy, and further release economic benefits.

IRDH275B-435  Third, large models will be fully perceptive with the help of multimodal technology. Zou Debao said that in the past year, the high-frequency words in the artificial intelligence industry are no more than the large model, as of December 2023, China has released as many as 234 large models above the scale of 1 billion parameters, and the number of large models and the scale of parameters have shown exponential growth. In the future, with the continuous development of multi-modal data processing and the integration of multi-modal perception and other technologies, large models will also continue to transform and integrate from the current single models such as text, pictures, audio and video to multi-models, and large models will realize comprehensive perception with the help of multi-modal technology.

Fourth, AI Agent (AI agent) will reconstruct human-computer interaction mode. AI Agent has been initially applied in task solving, service creation, intelligent customer service and other scenarios. However, with the development of interface alignment, complex task planning, working memory and other technologies, AI Agent application scenarios will continue to expand, and human-computer interaction will shift from traditional graphical interfaces to more natural human-IRDH275B-435  computer interaction. At that time, the AI Agent will reconstruct the human-computer interaction.

Fifth, embodied intelligence will be the next wave of disruption in manufacturing. Embodied intelligence is integrated with the hardware and entity of the physical world, integrating multi-disciplinary technologies such as sensing, machine vision, robot operation, intelligent control, wireless communication, and the Internet of Things, with the ability to perceive the external environment, make independent decisions and actions. With the iteration and maturity of large models in semantic understanding, visual perception and logical reasoning, embodied intelligence will make further breakthroughs in perception, reasoning and generalization capabilities, and embodied intelligence will promote manufacturing industry to achieve leapfrog improvement in production efficiency, quality control and cost savings.

IRDH275B-435  Sixth, the intelligent computing center will connect the points into a plane, and the artificial intelligence computing power network will coordinate and accelerate the deployment. The construction of large-scale intelligent computing centers has become a key technical element for the AI empowering industry and high-quality development. In the future, intelligent computing centers already deployed in various regions will be organically connected to build a new computing power network, realize the perception, distribution and scheduling of artificial intelligence across regions, and improve high-performance computing power to support the development of the AI industry. We will further promote high-quality development of the regional economy.

IRDH275B-435  Seventh, industry applications will accelerate the native transformation of AI, and the scene enables continuous innovation. At present, the industry can not meet the single deployment and operation and maintenance relying on the cloud computing platform, but must actively integrate the core logic of artificial intelligence at the beginning of the design. Zou Debao predicted that with the wide application and trial and error of artificial intelligence technology in various industries, as well as the close cooperation between enterprises and cloud service providers in generative AI platforms and development tools, more and more enterprises will complete the upgrading and transformation from cloud native to AI native in the future.

IRDH275B-435  Eighth, AI talents will jump from “algorithm type” to “composite application type”. Talent is one of the core elements of the whole cycle of artificial intelligence industry iteration, product innovation and scene application. In the future, with the continuous iteration and development of artificial intelligence technology, the coupling degree of artificial intelligence scene will continue to deepen, and artificial intelligence talent will also shift from a single algorithm type to a composite application type. Composite application-oriented talents such as cross-field talents who can actively use AI large models to empower business, AI algorithm talents who can explore application scenarios, and management and decision-making talents who understand AI large models will become the main talent demand in the field of artificial intelligence.

Ninth, AI governance will create more credible and controllable industrial applications. Big models are disruptive technologies that raise concerns about the risks of artificial intelligence. With the promulgation of the interim measures for the management of generative artificial intelligence services, the state is also continuing to strengthen the construction of artificial intelligence security governance capabilities, and will accelerate the future promotion of AI governance based on corpus data blacklist, “code regulation code” algorithm supervision technology and ethical governance routes, driving the creation of more credible, controllable and interpretable AI applications.

Tenth, MaaS (Model as a service) will become the core of the construction of AI industry ecology. MaaS develops pre-trained models through large model technology and provides standardized API (Application Programming Interface) interfaces for algorithm portability. With the development of large models, MaaS ecology gradually takes shape. Upstream enterprises provide pre-trained models as the base of artificial intelligence technology, and downstream industries develop personalized application models based on them that are in line with their own industry production and application, thus transforming artificial intelligence from “manual workshops” to “standard factories”. Therefore, MaaS will become the core driving force for building an artificial intelligence industry ecosystem.