Zhu Xiaohu: AI To C will experience an explosion next year.

```

Will AI reshape user entry points, business models, and industry divisions?

On September 11, at the 2025 Bund Conference Industry Roundtable, Ant Group CEO Han Xinyi, Xiaomi Group Mobile Department Vice President Zhang Lei, and GSR Ventures Managing Partner Zhu Xiaohu had a discussion on this hot industry topic. The leading figures from three key fields gave crucial insights respectively from the perspectives of applications, hardware, and investment.

From the wave of applications launched by Open AI, to the sudden rise of DeepSeek, the popularization of large models is also stirring up an internal industry battle over entry points: “Will large models eat up all software?”

In response to this question, investor Zhu Xiaohu believes that low-code and no-code software will definitely be replaced by AI, especially editing- and collaboration-type software. “It’s not that they will disappear, but their demand will drop. If the user count drops by 10%, that impact is already huge. Collaboration software will still exist in the future market, but it will be much smaller.”

He also revealed that when facing AI products, the only metric investors care about is user retention. “From PC Internet to mobile Internet to AI, it’s the same: the user recall cost for AI products may be more than 10 times that of mobile Internet products. Whether retention is good shows whether these companies have potential for further development,” said Zhu Xiaohu.

Han Xinyi, on the other hand, thinks that specialized AI vertical applications have irreplaceable value that general large models cannot deliver. Taking Ant’s AI layout in the healthcare field as an example, health is an essential need for almost everyone, and it is trending toward high frequency. “If AI healthcare can be perfected, there will no longer be concerns about retention. Understanding patients, providing personalized services based on actual situations, and helping implement solutions—like a professional doctor—that is the ultimate goal of AI healthcare.”

At the same time, he also believes that AI cannot replace doctors in the short term. It should be a doctor’s assistant. The current optimal solution is “human-machine collaboration”: “We hope AI can help doctors diversify their roles, allowing them to focus more on research and on difficult diseases, while enabling many grassroots doctors to have good assistants.”

In addition to mobile applications, AI today is almost “redoing” all hardware—AI phones, AI glasses, and AI robots are emerging one after another. Xiaomi Group Mobile Department Vice President and Wearables General Manager Zhang Lei believes, “The entry point opportunities of the future will still depend on whether the scenarios enabled by AI-empowered hardware have super stickiness. In the long run, AI glasses have great potential.”

“Compared to AI phones, AI glasses change the fundamental logic of interaction, moving from passive to active, potentially bringing about a change in external interactive paradigms,” Zhang Lei revealed. Currently, the frequency of calling “Xiao Ai” on Xiaomi’s glasses is already 6-7 times that of the phone end.

“In the AI field, next year there will definitely be an application boom. The next Bytedance, the next Xiaohongshu, should have already been founded this year. Entrepreneurs have lots of opportunities—everyone must have the courage to explore the vast universe.” When talking about AI business models and entrepreneurial opportunities, Zhu Xiaohu encouraged young entrepreneurs at the event.

Comparing fast-growing AI companies in China and the US, Zhu Xiaohu said the US focuses more on B-end entrepreneurship, while China excels at C-end. “AI is the same for everyone; the differences lie beyond AI itself. There are many ways to build differentiated user experiences outside AI, and this is what Chinese entrepreneurs are best at.” Han Xinyi and Zhang Lei both reviewed their respective practical experiences with Ant and Xiaomi and agreed that there’s still tremendous entrepreneurial space in vertical specialist service fields.

When asked about exploring business models in vertical scenarios, Han Xinyi gave an example of Ant’s AI healthcare. He said frankly, “For us, in the next one to two years, attention should not be on commercialization, because with a huge market and clear pathways, there’s no need to spend a lot of energy discussing it. Instead, emphasis should be on the three key challenges mentioned earlier: high-quality data chains, professional competence, suppressing ‘hallucination,’ and relatively complete medical ethics.”

Regarding the overseas expansion of AI hardware, Zhang Lei believes, “We may have cost and efficiency advantages in the supply chain, as well as a strong AI ecosystem advantage. But the key lies in finding truly necessary user scenarios within this excellent AI ecosystem and truly landing them to improve user experience, stickiness, and product strength. All this needs to be fed back to the product itself based on these advantages.”

Zhu Xiaohu also believes that all entrepreneurship must return to human nature: “Human needs are unchanged; they haven’t changed in the last 30 years, nor will they in the next 30. With AI, there will be better experiences and better product forms.”

Risk Disclosure and DisclaimerThe market has risks, and investment should be cautious. This article does not constitute personal investment advice, nor does it consider the special investment objectives, financial situation, or needs of individual users. Users should consider whether any opinions, views, or conclusions in this article are appropriate for their specific circumstances. Acting on this information is at your own risk. ```