Author: Zhou Yuan / Wall Street News
The current AI industry is shifting direction: the clamor of the large model parameter competition is fading, and the implementation of technology is beginning to become the new stage for the next round of industrial competition.
On May 21, OpenAI acquired the hardware company io, founded by Apple's former Chief Design Officer, for $6.5 billion, which is seen by the industry as a strategic turning point for AI's shift from &#34;cloud models&#34; to &#34;physical hardware.&#34;
This aligns with the five development stages of AGI (Generative AI) proposed by OpenAI: &#34;L1-L5&#34; (L1 Conversationalist → L2 Reasoner → L3 Actor → L4 Innovator → L5 Organizer); among them, the L3 stage requires the agent to achieve autonomous task execution, marking AI's leap from &#34;passive thinking&#34; to &#34;active action.&#34;
<img src="https://imageproxy.pbkrs.com/https://wpimg-wscn.awtmt.com/cedecd88-8c22-44cf-940d-3c429a84728b.png?x-oss-process=image/auto-orient,1/interlace,1/resize,w_1440,h_1440/quality,q_95/format,jpg" alt="" original-src="https://imageproxy.pbkrs.com/https://wpimg-wscn.awtmt.com/cedecd88-8c22-44cf-940d-3c429a84728b.png"/>
How do agents and hardware constitute the pillars of AGI technology implementation?
On June 19, at the Mobile World Congress (MWC Shanghai), Honor CEO Li Jian proposed a systematic technological framework.
Li Jian clearly pointed out that the ability to implement AI technology has become the core of industry competition.
Li Jian emphasized that when model capabilities converge, the value of AI is no longer limited to parameter competition, but must become a life partner that is &#34;right beside you,&#34; evolving from usable paths in the laboratory to a closed-loop experience that users are willing to use and frequently use in multiple scenarios.
Honor's thoughts and actions resonate with the three major transformations in the AI industry: the focus of competition is shifting from model performance to implementation capability, the value logic is shifting from tool efficiency to result closure, and the product form is shifting from cloud computing to personal hardware.
It is worth noting that the new generation of Honor's foldable smartphone &#34;Magic V5,&#34; set to be released on July 2, is the core carrier of this strategy, which Li Jian calls the &#34;strongest AI agent smartphone&#34; in the industry: through the ability to complete a closed loop with three actions—&#34;checking data, making PPTs, and transferring files&#34;—it becomes a golden key to solving the three major industry challenges of trust, performance, and scenarios.
<h2>AI Development Faces Three Major Transformations</h2>
From the overall context of technological industrial development, AI is undergoing three major transformations: the focus of competition is shifting from model performance to implementation capability, the value logic is shifting from tool efficiency to result closure, and the product form is shifting from cloud computing to a user's personal presence.
<img src="https://imageproxy.pbkrs.com/https://wpimg-wscn.awtmt.com/0a913331-bd6c-451d-a265-0041bb893faa.png?x-oss-process=image/auto-orient,1/interlace,1/resize,w_1440,h_1440/quality,q_95/format,jpg" alt="" original-src="https://imageproxy.pbkrs.com/https://wpimg-wscn.awtmt.com/0a913331-bd6c-451d-a265-0041bb893faa.png"/>
In the past, the AI industry often focused on the advancement of model parameters and algorithms; now, with the rapid development of AI technology, model capabilities are increasingly converging, making it a new focal point for enterprises to effectively integrate AI into daily life and realize practical value.
Previously, companies often competed on how much data AI models could process and how high their accuracy could reach; now, they are more concerned with whether AI can provide users with convenient, efficient, and smooth services in scenarios such as smart homes and smart offices
At the same time, AI is no longer just a tool for improving efficiency, but aims to achieve a complete closed loop from task initiation, execution to completion, bringing perceivable results to users.
Taking the office scenario as an example, in the past, AI might have only assisted in document writing, but now it needs to provide a one-stop service from data collection, content creation to document sharing, ensuring smooth connections at every stage, saving users time and improving work quality.
More importantly, AI is no longer limited to complex cloud computing, but is entering users' lives through various smart hardware, such as smartphones, tablets, smart speakers, etc., becoming an intelligent assistant by users' side.
For instance, smartwatches can monitor health data in real-time and remind users to exercise, while smart speakers can play music and query information based on user commands, making AI services easily accessible.
In this process, the agent becomes the key entry point for understanding user needs and executing tasks, and hardware has transformed from merely carrying model functions to being an extension of AI's &#34;personality,&#34; helping AI evolve from a cloud tool to a partner by users' side.
The release of AI's potential in the future depends on whether this technological capability can be integrated into life in a natural, seamless, and smooth manner; to achieve a close presence in all scenarios of life, it cannot rely solely on smartphones as a single terminal, but must be replaced by multiple terminal forms, allowing users to feel the value of AI in their daily use.
As a smartphone terminal company, Honor's move at the MWC Barcelona conference in March 2025 closely aligns with the transformation of AI's all-scenario demand for &#34;multi-terminal&#34; — Li Jian proposed the core of the &#34;Alpha Strategy,&#34; which is to promote Honor's transformation into a global AI terminal ecosystem company.
This time, Li Jian's explanation at MWC 2025 Shanghai can be seen as a deepening of Honor's new strategy in terms of specific paths.
Li Jian said, &#34;True AI applications are not about what it does after you click, but what it actively completes for you.&#34;
This viewpoint resonates with the shift in the AI industry towards application landing.
As technology thinker Kevin Kelly stated in an interview, &#34;The technium will evolve autonomously like an ecosystem, ultimately becoming the 'new nature' of human existence.&#34;
Based on this statement, a further inference can be made about Kevin Kelly's thoughts: &#34;The endpoint of technological evolution is biological; it will become so natural that it becomes part of our environment.&#34;
Honor's vision is precisely to hope that AI technology can seamlessly integrate into users' life scenarios in such a natural and smooth manner.
<h2>How to Break Through the Three Closed Loops?</h2>
Li Jian mentioned in his speech that for AI to truly enter users' lives, it must break through the three closed loops of &#34;scenarios, trust, and performance,&#34; which is key to solving the transition from &#34;technology being usable&#34; to &#34;users willing to use and frequently using.&#34;
The ultimate mission of AI is not merely to showcase intelligence, but to enable users to live more freely and conveniently; when technology integrates into daily life, users regain control over their lives, and the AI experience closed loop can be considered achieved.
From the perspective of industrial collaboration, Honor is committed to breaking down the &#34;three islands,&#34; initiating joint cooperation, and thus breaking through the three closed loops, deeply integrating into users' lives
<img src="https://imageproxy.pbkrs.com/https://wpimg-wscn.awtmt.com/7f95d88d-537d-4e07-a3ca-d65b95b35937.png?x-oss-process=image/auto-orient,1/interlace,1/resize,w_1440,h_1440/quality,q_95/format,jpg" alt="" original-src="https://imageproxy.pbkrs.com/https://wpimg-wscn.awtmt.com/7f95d88d-537d-4e07-a3ca-d65b95b35937.png"/>
First is the scenario closed loop, which connects islands and provides services on the go; the goal of the scenario closed loop is to allow services to revolve around users rather than users searching for services across different devices.
Currently, users have multiple devices, but the service experience is often fragmented. AI needs to understand user intent and seamlessly connect tasks across different devices.
Previously at MWC 2025 in Barcelona, Honor showcased its full-ecosystem file-sharing technology, achieving information (files, photos, etc.) transfer between Android, HarmonyOS, and iOS.
After breaking through the scenario closed loop, AI can &#34;do it well (cover the entire process), do it right (provide high-quality service), and do it all (cross-scenario collaboration)&#34; in users' life and work scenarios, allowing users to gain a sense of &#34;freedom&#34;—not having to worry about the fragmented and incoherent nature of AI services, thus focusing solely on achieving their own goals.
Secondly, the trust closed loop aims for safety and controllability, allowing users to use it with peace of mind.
Currently, AI faces many core issues, mainly including: generated content may contain factual errors (such as fabricating knowledge data); data collection may pose privacy leakage risks; and decision-making processes may touch on ethical boundaries.
Breaking through the trust closed loop requires addressing three major challenges in AI development, achieving technological co-creation, standard co-construction, and ethical co-governance, to provide users with a sense of &#34;security.&#34;
Computer science pioneer Alan Kay once said, &#34;The best way to predict the future is to invent it, and a trustworthy system must allow users to audit its decision logic.&#34;
The NSF (National Science Foundation) further elaborated on Alan Kay's quote: inventing the future must be based on a trustworthy architecture.
Building user trust in AI technology is an indispensable part of creating this &#34;trustworthy future.&#34;
When AI technology can continuously provide users with &#34;freedom (seamless service), security (trustworthy and controllable), and satisfaction (strong and smooth),&#34; it can truly become an omnipresent intelligent partner in users' lives.
This is also the underlying logic for companies represented by Honor to build an AI ecosystem: guided by user experience, breaking through three closed loops, and dismantling system barriers between enterprises, thereby promoting the AI industry from &#34;technological competition&#34; to &#34;value inclusiveness.&#34;
Third, the performance closed loop aims for a smooth experience, with the method being on-demand collaboration.
Powerful AI functions require sufficient computing power support, but the computing power of terminal devices always has its limits. The performance closed loop emphasizes intelligently allocating computing resources through end (device side) and cloud (cloud side) collaboration based on task requirements and device status, ensuring users receive smooth and timely responses, avoiding stuttering or delays.
Honor's &#34;multi-agent collaboration&#34; technology is based on this concept, allowing multiple AI agents to collaborate to complete tasks, such as intelligently calling local and cloud resources when generating complex documents, thus making AI &#34;powerful, smooth, and efficient&#34; to meet user needs
<h2>The Key to Ecological Co-construction</h2>
To break through the three major closed loops and accelerate the implementation of AI, it is undoubtedly difficult to achieve this solely through the efforts of one company.
Therefore, Honor calls for the establishment of an open AI terminal ecological alliance based on the GSMA platform: uniting AI model companies, operators, AI terminal companies, and internet companies to form a close alliance, complement each other's strengths, and cooperate openly. Together, they will engage in technological innovation, business model innovation, standard definition, and talent cultivation, jointly leading industry transformation and accelerating AI implementation.
<img src="https://imageproxy.pbkrs.com/https://wpimg-wscn.awtmt.com/276178ed-3276-4b5f-81c0-bcc008d8e52c.png?x-oss-process=image/auto-orient,1/interlace,1/resize,w_1440,h_1440/quality,q_95/format,jpg" alt="" original-src="https://imageproxy.pbkrs.com/https://wpimg-wscn.awtmt.com/276178ed-3276-4b5f-81c0-bcc008d8e52c.png"/>
In AI ecological construction, Huawei collaborates with chip companies and application developers to jointly build the Harmony AI ecosystem, enhancing the operational efficiency and user experience of AI in terminal devices through the synergy of systems, chips, and applications. Google, through the Android system, opens AI capabilities to global developers, promoting the diversified development of AI applications.
The AI terminal ecological alliance proposed by Honor aligns with industry trends, integrates resources from all parties, and promotes the widespread application and in-depth development of AI in terminal devices.
At the Shanghai World Mobile Communication Conference, Honor announced in collaboration with China Mobile that both parties will focus on AI large models and intelligent agent technology as core driving forces to initiate comprehensive and in-depth innovation cooperation in terminal products.
This action by both parties marks the beginning of in-depth cooperation between the largest mobile communication operator in terms of global network scale and customer scale and a leading global AI terminal ecological company, relying on their respective core advantages to focus on product innovation, ecological co-construction, and the implementation of business models for AI terminals. This can be seen as a model for cooperation within the AI terminal ecological alliance.
Overall, Honor's layout in the AI field can be described as proactive and positive: from proposing the Alpha strategy to advocating for the establishment of the AI terminal ecological alliance, from showcasing the Magic V5 features to emphasizing the breaking of three closed loops, closely following the pulse of industry development.
At this critical stage where AI moves from &#34;thinking&#34; to &#34;action,&#34; Honor takes a proactive stance to lead industry transformation, bringing AI into the everyday life scenarios of users, providing intelligent, convenient, and safe experiences.
This series of initiatives by Honor injects new technological momentum and industrial vitality into industry development and outlines a value vision blueprint for the future of AI.
The potential of AI technology is enormous, but the industrial value it releases ultimately depends on whether AI technology can provide practical and natural assistance in every breath and every life scenario of users.
The technical and ecological conceptual framework proposed by Honor is not only a phase of reflection facing the dramatic changes in the AI industry but also a rare practice worth advocating to break down corporate interest barriers.
As the saying goes, &#34;A tree that is a hundred feet tall grows from a tiny sprout; a nine-story platform rises from a pile of earth.&#34;
The widespread application of AI technology requires grounded technological advancement and gradual accumulation.
This also requires companies like Honor to continuously invest resources, unite industry partners, and always focus on real user needs, engaging in continuous exploration and optimization

GOOGL

GOOG

OpenAI

HUAWEI

AI industry upheaval: Li Jian's countermeasures

- The AI industry is shifting focus from model parameters to practical applications, highlighted by OpenAI's acquisition of a hardware company, marking a strategic pivot towards physical hardware.  
- Honor's CEO emphasized the importance of AI's integration into daily life, advocating for a seamless user experience across multiple devices and addressing challenges in trust, performance, and usability.  
- The proposed AI terminal ecosystem alliance aims to foster collaboration among industry players to enhance AI's practical value, ensuring it becomes an integral part of users' lives.

Track Hyper | The Sword of Glory Aims at New Competition for AI Capability Implementation