aifollow.news
返回 IT之家 科技新闻
IT之家 科技新闻· · 原发布时间 AI 评分27

杨元庆提出“私域数据飞轮”,称 AI 竞争将转向混合式 AI 落地

AI 辅助摘要

联想集团董事长兼 CEO 杨元庆认为,个人和企业的数据可用于训练智能体,而智能体越懂用户,使用越多,就会产生更多数据。他称 AI 产业竞争将从模型参数转向混合式 AI 的落地能力。演讲还披露,联想天禧个人超级智能体已升级至 4.5。

正文 · AI 翻译

IT Home reported on November 16 that Lenovo Capital's 2026 CEO Annual Conference, themed “Gathering Moments Together, Advancing Technology Toward the New,” was held in Hangzhou, Zhejiang, from November 15 to 16.

Lenovo Group Chairman and CEO Yang Yuanqing delivered a keynote speech titled “Accelerating Transformation with Hybrid AI to Create a Future of AI for All,” explaining Lenovo Group's latest assessment of AI industry trends and the progress in implementing its hybrid AI strategy.

Yang Yuanqing said in his speech that AI is moving from generation to execution, and agents are becoming the core of value creation. He said AI's success should be measured not only by how many devices have been sold or how much infrastructure has been built, but also by whether agents are being used and continuously creating value. For individuals, a successful agent must truly understand you and ultimately become your AI twin; for enterprises, a successful agent must be able to cross data domains and break down application silos, allowing people and agents to work side by side. How deeply an agent understands depends on how rich the data is and how deep and long-lasting its memory is.

He proposed a “private-domain data flywheel”: personal devices and enterprise operations generate data, which trains agents; the better agents understand users and enterprises, the more willing users and enterprises are to use them, generating more data. Everyone has access to public data, and models will become increasingly similar. What truly sets them apart is unlocking the potential of each user's unique private-domain data through personal and enterprise intelligence.

Yang Yuanqing predicted that competition in the AI industry will shift from comparing model parameters to comparing the ability to put hybrid AI into practice. “The next battle in AI will depend not only on models, but also on data and hybrid AI.” He stressed that data accumulated by enterprises across various applications belongs to the users themselves, not to application software or SaaS providers.

In personal intelligence, Lenovo is advancing its “one platform, multiple devices” strategy. In the third quarter of this year, Lenovo's global AI PC market share reached 25%, keeping it in first place. In China, AI PCs accounted for more than 50% of Lenovo's laptop sales, with cumulative sales exceeding 7 million units to date. Lenovo continues to expand its range of devices, including AI desktop computers and AI glasses. A total of 42 AI devices passed the first round of L3-level testing, the highest number in the industry; the Tianxi personal super agent has been upgraded to version 4.5, moving from responding to requests to autonomous execution.

In enterprise intelligence, Lenovo continues to strengthen its “device-edge-cloud” hybrid AI infrastructure and accelerate its development of AI inference computing and edge computing. Yang Yuanqing believes that, in the future, 70% to 80% of tokens will be generated at the edge or on devices, and 20% to 30% in the cloud.

IT Home noted that this year marks the 10th anniversary of Lenovo Capital's founding. Yang Yuanqing said that over the past 10 years, Lenovo Capital has invested in more than 350 promising technology companies, 27 of which have gone public through IPOs. More than 70 portfolio companies are currently collaborating with Lenovo on business around its hybrid AI architecture, spanning computing infrastructure, models, applications and other areas. In the past year alone, the value of ecosystem collaboration between Lenovo Group and its portfolio companies exceeded RMB 10 billion.

发现内容有误?提交纠错