[Abstract] As a leading player in the PC era, Lenovo, rooted in "trade-industrial-technology," now stands at a pivotal juncture in the AI era, confronting a critical dilemma.

From investing $1.2 billion in a comprehensive AI strategy to pioneering AI PCs but grappling with technical scrutiny, Lenovo's AI approach seems bold yet harbors underlying concerns.

As Huawei and Apple fortify technical barriers with ecosystems and chips, Lenovo must decide whether to persist with "assembly thinking" through supply chain integration for swift market positioning or pivot towards investing in core technologies. Lenovo must reevaluate its future strategy.

Below is the main text:

AI Journey: Seven Years of Wins and Losses

Amidst sluggish global PC sales growth, Lenovo chose to invest over $1.2 billion in artificial intelligence, the Internet of Things, and big data in 2017. Since then, the company's focus has shifted from selling smart devices to intelligent agents.

At the time, Yang Yuanqing, Chairman and CEO of Lenovo Group, asserted that AI is the future of the information industry, and the company must commit fully.

Since the introduction of ChatGPT in 2022, large models have become the focal point of industrial competition. Various AI agents are accelerating their application scenario expansion, and spending in the AI sector is escalating.

On its path to hybrid AI transformation, Lenovo embraces a full-stack AI approach, encompassing AI terminals, AI infrastructure, and AI solutions and services.

Among these, Lenovo's AI terminal strategy focuses on creating intelligent agents that can autonomously perceive the environment, make decisions, and execute actions, primarily leveraging the company's PC and other intelligent hardware businesses.

Under the "one body, multiple terminals" strategy, Lenovo has launched Tianxi Personal Super Intelligent Agent, the industry's first enterprise super intelligent agent Lenovo LeXiang, and AI PCs, AI phones, and AI tablets equipped with Tianxi Personal Super Intelligent Agent.

Moreover, the company pioneered the world's first AI PC last May and released an AI workstation integrated with the DeepSeek large model this spring, featuring the world's first AI PC deployed with a 7 billion-parameter DeepSeek large model on the edge side.

However, despite its aggressive strategy and diversified approach, Lenovo faces a market that has yet to coalesce around a unified definition.

For instance, Apple defined smartphones, and Tesla defined electric vehicles, both driving intelligent hardware iterations based on specific standards. This benchmarking strategy granted Apple and Tesla a first-mover advantage.

Conversely, Lenovo, fiercely competing in the AI industry, has yet to establish industry recognition through its diversification strategy. Public information suggests that Lenovo is in a situation where diverse products proliferate, but killer products lack sufficient impact.

Certainly, this radical decision is intertwined with Lenovo's historical context.

In the early stages of its strategic transformation, the company temporarily lost its position as the top PC shipper, falling behind HP. Last year, Lenovo's PC shipments reached 61.8 million units, and its market share returned to 23.5%, consistently ranking first.

Being at the pinnacle is isolating. Lenovo, which has long dominated the PC shipment throne, inevitably faces a higher opportunity cost in entering the AI race and is understandably more vigilant about its market position.

But deeper still, this reflects Lenovo's inheritance of its "trade-industrial-technology" roots.

In the 1990s, Lenovo experienced the "Liu-Ni Debate." Ni Guangnan, who insisted on research and development as the foundation and using labor achievements to drive production and trade, was defeated by the market faction in 1995 and eventually left in disgrace.

From acquiring IBM's PC business to integrating Motorola mobile phones and Fujitsu of Japan, Lenovo's path to scale expansion has always leaned towards capital operations.

According to official data from Lenovo Capital, as of now, the company has invested in over 50 AI-related enterprises, indicating that on this new trajectory, the company still opts for an external absorption commercial approach.

This thinking has yet to yield significant results in the AI era.

"New Bottle, Old Wine" for Intelligent Agents

From a technical standpoint, as an AI infrastructure provider, Lenovo's AI PCs still rely on Intel, Qualcomm, and AMD chip platforms.

Apart from outsourcing chips, the company's supplier list also includes Microsoft for the operating system, Samsung for hard disk memory, Foxconn for components, and AU Optronics for display panels.

Strong supply chain integration capabilities are indeed one of Lenovo's core competencies. Collaborating with leading manufacturers in various fields allows the company to minimize costs based on each company's strengths, thereby enhancing gross profit margins. However, the technological core in hand is often the linchpin to AI era success.

Notably, Lenovo has also chosen to integrate external technology on the model side, using an alliance approach to mitigate the high risks and long cycles of independent research and development.

At the end of 2024, Lenovo initiated the establishment of the Intelligent Agent Network Alliance, partnering with 20 industry giants such as Baidu, Alibaba, Volcano Engine, and Zhipu to jointly launch the AI Terminal Intelligent Agent Innovation Professional Committee of the China Software Industry Association.

Currently, the company's core technology for super intelligent agents heavily relies on third-party large model interfaces like DeepSeek. However, as DeepSeek connects with numerous vendors, Lenovo's advantages will further diminish.

This construction method renders Lenovo's intelligent agents highly similar in functionality to large models from companies like Baidu and ByteDance, lacking differentiated competitiveness. Overall, while this strategy can compensate for capability shortcomings, the digestion and transplantation of technology routes often require a longer time frame.

From an industry perspective, Canalys report data indicates that global AI PC shipments will reach 205 million units by 2028, with a compound annual growth rate of 44% from 2024 to 2028.

Market opportunities abound. For AI PCs to thrive, the key lies in penetrating the high-end market, and the prerequisite for entering this market is technical distinctiveness.

Compared to peers, in the domestic market, Huawei has been promoting the HarmonyOS ecosystem, supported by the Kirin PC processor and Kirin 9000S mobile processor. In the overseas market, Apple, Microsoft, and OpenAI collaborate to build competitiveness in the AI market.

In fact, as of the end of 2024, Lenovo has been listed in a total of 162 scientific computing clusters in the global TOP500 supercomputer (HPC) rankings, with a global market share of 32.4%, ranking first for the thirteenth consecutive year.

This amply demonstrates that the company possesses innovative capital, but its resources are not concentrated. Crafting its unique features is an issue Lenovo must contemplate next.

Epilogue: An Unfinished "Fifth" Entrepreneurial Journey

As Yang Yuanqing stated in an interview with Caixin, transformation is not an easy journey for any company. If it doesn't peel off several layers of skin, it's not a genuine transformation.

Currently, the AI strategy has become a ubiquitous choice in the technology industry, but without a profound definition, building product strength is challenging, and the industry's development is far from reaching the final stage of diversified competition.

Lenovo's fifth decade is the era of artificial intelligence and the company's fifth entrepreneurial journey. Success or failure is both a short-term choice and a long-term plan.

Lenovo must reevaluate its future strategy.

- XINLIU -