new economy

　　

Xinhua News Agency, Las Vegas, USA, January 9th - International Observation | New Trends in Artificial Intelligence at the 2026 US Consumer Electronics Show

Reporter: Tan Jingjing, Huang Heng

The 2026 US Las Vegas Consumer Electronics Show concluded on the 9th. At this exhibition, artificial intelligence (AI) remained the most prominent keyword. However, unlike previous years, this year's focus has shifted from "model capabilities" to "product capabilities", and from a technical showcase to scalable applications. Journalists observed on the spot that AI is moving from conceptualization to engineering and industrialization, with edge AI and physical AI becoming the core directions of technological breakthroughs.

AI becomes "system-level capability"

Entering the CES exhibition halls, a notable change is that discussions around AI are no longer concentrated on model scale, computing power indicators, or technical route disputes, but more on what AI "can do", "how to integrate into existing systems", and "how to achieve large-scale deployment".

At the exhibition, large models remain the foundation, but they are no longer at the center of the exhibition booths. Instead, there are a large number of practical AI devices and system solutions that can be actually operated. Autonomous decision-making AI agents, physical AI systems that can interact with the real environment, and AI factories for enterprises and industries have become the focus of attention.

When communicating with several exhibiting companies, the reporter found that enterprises no longer repeatedly emphasized model parameters or computing power scale, but directly demonstrated how AI is embedded in devices, systems, and real scenarios. AI is becoming the underlying capability that is "default" in products, rather than a separately emphasized technical highlight.

In the personal terminal exhibition area, multiple manufacturers showcased computers, smart glasses, and wearable devices with local AI processing capabilities; in the automotive exhibition area, AI was integrated into the vehicle electronic architecture, forming a unified system from cabin interaction to assisted driving; in the industrial scene exhibition area, AI was applied to production systems and management platforms for production scheduling, quality inspection, and equipment maintenance.

Overall, CES is shifting from a "frontier technology showcase" to an "industry implementation test field". AI is deeply integrated into operating systems, vehicle platforms, and industrial control systems, becoming the fundamental capability that reshapes product forms and industrial processes.

Physical AI accelerates breakthroughs

Robots are one of the most highly attended exhibition areas at this year's CES. From humanoid robots, autonomous mobile robots to service robots and industrial robots, the exhibited products cover various scenarios such as manufacturing, logistics, retail, and elderly care. The focus of the display is no longer on "whether it can move", but on "whether it can operate stably for a long time".

Unlike previous years where the focus was on concept demonstrations, this year's robot exhibits are clearly targeted at specific application scenarios and place greater emphasis on practical application capabilities. The reporter observed in the exhibition hall that multiple robots were operating continuously in simulated warehousing, retail, and public service environments, capable of autonomously planning routes, identifying items, and completing multiple-step operations.

In logistics and warehousing scenarios, robots undertake tasks such as sorting, handling, and replenishing shelves; in manufacturing scenarios, they perform high-repetition or high-risk processes; in service scenarios, robots take on roles such as tour guides, delivery personnel, baristas, and tea artists, providing emotional value and stable, meticulous service experiences.

Cliff Henderson, the dean of the College of Engineering at the University of Alabama in the United States and a professor of chemical and biological engineering， this year's CES robot displays are rich in variety and more powerful in function. Especially， service robots have shown outstanding performance in automation and stability. He particularly pointed out that China's robot development is rapid and can continuously and stably perform daily tasks， with robust and reliable performance. With the continuous progress of perception， computing， and control technologies， robots are gradually moving out of the laboratory and entering real production and life scenarios.

The edge and the cloud are collaborating

At this year's CES, there has been a significant change in the selection of where AI is run. More and more products are choosing to run AI models locally on terminal devices to meet requirements for real-time performance, energy efficiency, and privacy protection, while the cloud continues to handle training, updates, and cross-device collaboration functions.

This trend is particularly evident in areas such as personal computers, vehicle systems, and wearable devices. Some exhibits demonstrated the ability to complete complex reasoning and interaction in an offline state, reducing reliance on the cloud.

Meng Pu, the chairman of Qualcomm's China region,在接受新华社记者采访时表示， the future of AI is not a simple choice between "cloud" or "edge". The model capabilities of the cloud and the immediate response capabilities of the terminals need to complement each other and evolve collaboratively. The collaboration of the edge and the cloud as the division of labor has become an important architectural feature of future AI development.

CES reflects changes in industrial logic

From this year's CES， it can be seen that AI is not only changing product forms but also reshaping industrial divisions. As AI accelerates its implementation， the importance of fields such as chips， manufacturing， and system integration has significantly increased. AI is no longer just the "home ground" of internet enterprises.

At the same time， deep collaboration between software and hardware has become a competitive focus. A single technological breakthrough is no longer able to form a sustained advantage. Complete system capabilities and large-scale delivery capabilities are becoming new thresholds. Industry insiders point out that AI is moving from the "rapid trial and error" stage of technological innovation to a stage that places greater emphasis on reliability， security， and long-term operation.

From this perspective， CES is no longer just a showcase window for new consumer electronics products； it is also becoming an important platform for observing the global AI industry's direction. The transformation from concept to implementation， from virtual to physical， is accelerating on the CES stage.