Entrepreneurs become bloggers, and investors build personal IPs: There is no "behind the scenes" in AI venture capital.
In 2025, the focus of the venture capital circle still remains on the fields of AI and embodied intelligence.
In the last few days of the year, the AI agent company Manus was acquired by Meta for billions of dollars. Among the "Six AI Stars", Zhipu AI and MiniMax went public in Hong Kong one after another, with their market values reaching tens of billions.
The financing rounds of leading companies have continuously set new records, pushing the industry's popularity to new heights. While capital, technology, and talent are flooding in, a notable change is taking place among entrepreneurs in the AI era: they no longer hide behind codes and papers but choose to step onto the stage actively.
Manus's sudden rise can be seen as the starting point for social media traffic to change the AI startup ecosystem this year. At that time, the startup team became an overnight sensation across the internet with a product introduction video. Within a day, a single invitation code was resold for up to 100,000 yuan. Even after the parent company's influence spread overseas and the content of its Chinese accounts was cleared, co-founder Zhang Tao has been posting updates on the product's business progress through his personal social media to maintain the hype.
At the end of this year, the "AMA" (Ask Me Anything) topic on Xiaohongshu, which has received over 300 million views, also originated from entrepreneurs. It started when Xu Huazhe, an assistant professor at Tsinghua University and co-founder of Xinghai Map, posted on Xiaohongshu and quickly spread to the venture capital circle of AI and embodied intelligence. A group of entrepreneurs who used to stay in the background, such as Yin Qi and Wu Weijie, as well as team members from companies like Zhipu AI, Qianxun Intelligence, and Minimax, have started to participate. They have joined Xiaohongshu in their personal capacities, frequently initiating AMAs and sharing their thoughts on the industry.
From competing on technology to competing on narrative
Reporters from the Science and Technology Innovation Board Daily noticed that at least four of the "Six AI Stars" - the core members of Moonlit Side, Zhipu AI, MiniMax, and Zero One Universe - are actively building their personal IPs on platforms like Xiaohongshu. Meanwhile, top technologists in cutting - edge fields such as embodied intelligence have also joined in. For example, Ding Wenchao, the co - chief scientist of Tashii and a doctoral supervisor at Fudan University, and Yu Yinan, the founder of Vita Power, have communicated with the public about technical difficulties through AMAs.
This is not just a simple expression of personal desire. Behind it is a new reality woven by capital and technology. Since 2025, the financing amounts completed by leading companies have reached hundreds of millions or even exceeded one billion. The barriers to technological competition have been set extremely high. However, having large amounts of capital also brings new challenges: how to explain the value of complex large - models or embodied intelligence to a market that is not proficient in technology?
An entrepreneur engaged in the research and development of embodied intelligence described their current situation to reporters from the Science and Technology Innovation Board Daily. "Most of the time, we are exploring an uncharted territory where even the 'problems' have not been clearly defined."
He gave an example. When developing a companion robot, the team initially assumed that the demand was "emotional companionship". However, after continuous conversations with potential users on social platforms, they found that the more urgent needs of elderly people living alone might be "fall prevention" and "physical assistance in emergencies". "Technology itself is iterating rapidly, but the real product definition emerges through interaction with users. If you don't step onto the stage, you may be solving a false proposition all the time."
As he said, whether it is AI or embodied intelligence, these are fields with extremely high technical thresholds but great market potential at the consumer end. However, their final forms and core values are extremely vague. This uncertainty makes "communication" an indispensable part of the productization process.
"Investors now explicitly encourage founders or core executives to step forward," the head of an AI company told the Science and Technology Innovation Board Daily. "In the current situation where there is no consensus on the AI valuation logic, an external image that can clearly explain the technology, define the track, and attract market attention can significantly reduce the 'explanation cost' for investment institutions and, to some extent, reduce the 'risk premium' they perceive."
From technological breakthroughs to product development, and then to obtaining market orders, each step involves a difficult "translation" - it is necessary to transform laboratory terms into business trends that investors can understand, specific values that users can perceive, and cooperation opportunities that partners can see.
It can be observed that CEOs or chief scientists are becoming the company's 'chief translators'. Scientists such as Xu Huazhe, co - founder of Xinghai Map, and Ding Wenchao of Tashii have shared complex concepts such as reinforcement learning and embodied intelligence with the consumer end through AMAs. They have also spread information about the company's recruitment, business progress, and technological iterations.
Meanwhile, this role is also becoming professionalized. The human resources director of a US - dollar fund confirmed this trend to the Science and Technology Innovation Board Daily: "We have received many recruitment requests from invested companies. In the past, they mainly focused on technical talents. Now, many companies are starting to add positions such as'social media brand 'technology communicator'."
Another comparable phenomenon is that Baichuan Intelligence, one of the "Six Large - Model Stars", has significantly lower activity and influence on public platforms compared to its peers. In terms of financing, public information shows that its recent financing rhythm and scale have been relatively stable. This difference may imply a divergence in strategic choices among different companies: in a situation where attention directly affects perception and resource allocation, remaining silent may mean actively giving up a crucial window for building user understanding and industry influence.
Overseas practices provide a reference. Aravind Srinivas, the founder of Perplexity, has been actively involved on the X platform. By continuously sharing product iterations and industry insights, he has not only accumulated early loyal users but also received valuable feedback for improvement. People like Sam Altman of OpenAI and Brett Adcock of Figure AI have closely linked their personal influence with the company's technological progress. Every opinion they share externally continuously shapes and guides market expectations.
Investors "step in"
Investors "entering the game" is another key signal at present. Since this year, a new generation of investors such as Yang Jie, the founding partner of Jinqiu Fund; Lei Ming, the founding partner of Sky Workshop Venture Capital Fund (formerly Zhuichuang Venture Capital); Yang Xuan of Xiaomi Strategic Investment; and He Zhiliang of ZhenFund have also started to turn their Xiaohongshu accounts into a combined business card for themselves and their institutions.
Yang Jie, the founding partner of Jinqiu Fund, previously worked at ByteDance. The fund she founded focuses on early - stage "AI - Native" investments, with a management scale of over $500 million. It has invested in 50 AI companies in a year. On Xiaohongshu, she systematically shares her investment methodology and insights into the industry, essentially defining the industry publicly and screening like - minded entrepreneurs. Lei Ming, the founding partner of Sky Workshop Venture Capital Fund, has started to use social media through AMAs to attract startup teams in the fields of AI and general robotics.
In the past, investors mainly played the role of "judges" sitting behind the negotiation table. However, today, with the rapid iteration of AI technology and the exploration of application scenarios, young investors are starting to enter the market at the same time as entrepreneurs.
"Almost every analyst born in the 1995s and 2000s in our fund has a Xiaohongshu account," the head of a dual - currency fund told reporters. "In the past, when we evaluated projects, we relied on recommendations from FAs (financial advisors) and industry summits. Now, how a team explains its technology on Xiaohongshu, how it interacts with comments, and even the founder's study notes have all become materials for our initial 'psychological profiling' and judgment of the team's temperament. In a sense, the starting point of due diligence (DD) has moved forward to social media."
According to her, they not only encourage young colleagues to keep in touch with the industry through Xiaohongshu but also recently started to officially operate an IP account, regarding it as a three - in - one platform for talent recruitment, project discovery, and brand building.
It is foreseeable that this change in the way of obtaining information will inevitably affect the project discovery mechanism in the future. In the past, project sources highly relied on FA recommendations and introductions from acquaintances, and the chain was relatively closed. Now, a technical geek's late - night post about a technical challenge or a R & D note of a consumer product may directly catch the attention of investors.
More importantly, this kind of observation is continuous and multi - dimensional. Another investment director who focuses on early - stage hard technology shared his new approach: "I will long - term follow the accounts of founders or scientists I am optimistic about. I observe how they continuously think about a technical problem, how they debate with peers, and even what booklists and insights they share in non - professional fields. These fragments can piece together a 'personality and cognitive profile' that is much more vivid and real than a resume and a business plan."
In a field like AI, where the technological prospects are broad but the commercialization path is unclear, investment is, to some extent, "investing in people". When entrepreneurs take to the new stage, investors follow.
This "front - stage" process is essentially a communication adjustment in the process of technology entering the market. It allows the voice of innovation to be heard more directly and extends the reach of capital to more real application scenarios.
However, it is worth noting that narrative ability itself is not the underlying value of a startup team. In technology - driven fields, the core barriers always come from technological innovation, engineering implementation, and product refinement. Public expression and personal brand building are more of auxiliary abilities to add icing on the cake.
In the long run, the market's final judgment will still be based on actual results rather than the volume of voice. Products that can continuously solve real problems and create user value are the key to surviving market cycles and gaining recognition. In this sense, "being seen" is only the starting point, and "being needed" is the end - goal.
This article is from the WeChat public account "Venture Capital Daily", author: Yu Shiqi. It is published by 36Kr with authorization.