AI Observation: 99% of tech founders actually do not need to build IP.

In an era where model advancements can overshadow all application innovations and models are updated on a monthly cadence, waiting until the product is fully ready to speak up often means you are already too late.

The logic of talking about technology and products seems to have lost its effectiveness. As a result, more and more Founders are stepping onto the front stage, turning themselves into the entry point for the market to understand their companies.

Sam Altman's Twitter account has practically become OpenAI's bulletin board, where he frequently explains when products will launch, why they are delayed, and where the company will head next.

Jen-Hsun Huang's first post on X was a joint letter signed by NVIDIA that discussed why open-weight AI is so important.

The same is true in China, especially for automotive companies, where the founder's image is inextricably tied to the car brand. Delivery figures, strategic goals, and product roadmaps are almost always first announced via personal accounts, followed by official corporate statements.

We can also see that when their products and models are still being refined, the founders of the "Six Dragons" have frequently participated in media interviews, delivering one memorable quote after another.

These public-facing CEOs bypass the traditional "company-PR-media-public" communication path, directly stepping into the information streams of users, developers, and investors to meet their psychological need to participate in corporate narratives, align with social media algorithms, and achieve high brand exposure.

If a Founder plans to step onto the front stage, do they need to build a personal IP? What do they need to deliver to the market?

The answer is probably not an account, not fans, not a public persona, not perfection, and not content, but something else entirely.

Not an Account: 99% of Founders Do Not Actually Need to Build a Personal IP

When Founders decide to speak out publicly, their first reaction is often to create an account first.

Short videos, podcasts, daily updates, personal branding — these quickly become standard practices. The market even has teams that specialize in helping company bosses plan topics, film, edit, and operate their accounts.

But Jiang Xu does not think this way. As the host of the podcasts "Just a Moment" and "When You Become Famous", who was recently interviewed by "UnDefined", she has met countless Founders, and her observation is: Most company bosses do not actually need their own personal accounts.

Especially for To B companies, bosses do not need to post content every day. Launching new products, securing financing, and strategic pivots — these are the moments when they should speak up. Even if they only speak publicly two or three times a year, it is far more useful than forcing out content just to maintain a posting frequency.

Therefore, what Founders need is the ability to express themselves, not necessarily to produce content.

But there are nuances to effective expression.

Many Founders are accustomed to the standard corporate PR process: find a media outlet, align the outline in advance, and speak according to pre-prepared answers.

This seems safe, complete, and error-free, but the statements produced are often unconvincing to both sides: industry insiders find them too superficial, while outsiders cannot understand them, and in the end no one remembers what was said.

What the public wants to hear is your real thoughts, not the standard answers written by the PR department.

Jiang Xu's suggestion is: appear on other people's shows more often. Answer their questions and share your real judgments. A single in-depth conversation is far more memorable than dozens of carefully edited short videos in letting people know who you are.

Not Fans: What Founders Truly Need is High Density Among Key Groups

When Founders speak out publicly, who exactly is the "public" they are addressing?

100,000 followers, or 1,000 people who can actually influence your business — which would you choose?

For technology companies, the latter is the correct answer.

A podcast episode with 5,000 plays is not a large number in the mass content ecosystem. But if among those 5,000 people there are CEOs, investors, engineers, candidates you want to recruit, and real decision-makers in the industry, that counts as a precise hit for a tech company.

What Founders truly need is to have enough key people think of them immediately when relevant needs arise.

Jiang Xu prioritizes the public influence of Founders in this order: First is talent recruitment, financing, and customer acquisition; second is industry influence and organizational cohesion; brand building comes last.

Starting with talent recruitment: in the AI industry, the final competition boils down to talent. When an engineer accepts an offer, they do not only care about salary and title — they also want to know where the company is heading, and whether they can grow professionally after joining.

The same applies to financing. AI companies face high levels of uncertainty, and investors cannot judge what the company will look like three years from now based solely on its current products. Therefore, how Founders view the future and how they will adjust to new changes becomes extremely important.

The same goes for clients. When placing their first order, they are not buying your current version of the product — they are betting that your company will still be around in a year, and that you will be able to keep up with technological changes in three to five years.

At the end of the day, talent recruitment, financing, and customer acquisition all boil down to the same core question: why should others trust you? Why should they work with you? Why should they give you their money and business?

In the past, all these questions were attributed to the quality of a company's brand.

But brand is actually a result. When people are willing to join your company, investors are willing to trust you, clients are willing to give you their business, and the industry is willing to listen to you, all of these accumulated over time gradually form a brand.

Therefore, the metric for measuring a Founder's public influence should be density of key groups, which directly determines where opportunities flow.

Not a Persona: A Public Personality Is Not Designed by the Boss Themselves

After Founders are seen by more people, they immediately face a problem: what kind of image do I want the outside world to remember about me?

"Public persona" seems to be the universal answer, but Jiang Xu is reluctant to use this term. She thinks it sounds like acting: you write a script first, then follow it to perform.

A person's public image is actually co-created by multiple parties.

You have real experiences and real capabilities, which form the foundation. The company selects some materials from your experiences to release to the public, media outlets pick the topics they find newsworthy from those materials, and in the end the public only remembers one or two labels about you.

Yang Zhilin is a very interesting case for observation.

Post-90s prodigy, management genius, recipient of the special Tsinghua University scholarship, co-first author of two top journal papers... these were originally unrelated events in his life.

But once a person becomes famous, society will retroactively re-edit their past, stitching together otherwise unrelated events into a "fated" narrative: he was like this from childhood, so he was destined to succeed later.

This shows that a public personality cannot be fabricated out of thin air by a PR team. What you end up being remembered for is not entirely under your control.

Instead of rushing to design a perfect role for yourself, you should first ask yourself: Is there anything real about me that can withstand public scrutiny?

Not Perfection: The Sense of Authority of "Always Being Right" Is Fading

If a personality cannot be designed, then as a compromise, can you package yourself to look perfect?

The kind of authority people used to believe in was defined by clear direction, unambiguous answers, rich experience, firm stance, and very few mistakes.

But the AI industry is changing so fast that this set of standards no longer works. A judgment that is correct today may be wrong just a few months later.

Therefore, the current sense of authority is shifting to the ability to continuously update your answers.

Yan Junjie has revised his own understanding before. When he was interviewed by LatePost the previous year, he was still thinking about growth using the DAU metrics from the mobile internet era; when they talked the next year, he said that approach should never be used anymore.

"Revising your views" does not equal "being fickle". To truly convince people, you have to clarify at least three things: Why I made that judgment at the time; what changes happened later; why I changed my mind today, and what actions this change ultimately affected.

If you only say "I don't think that way anymore", that is not iteration — it is just swapping out one answer for another.

Therefore, what truly convinces people is making your mistakes traceable and your revisions explainable.

This is not easy for many Founders. Many young, high-achieving entrepreneurs find it hard to accept being questioned, as the rules that have worked for them repeatedly in the past are to get things right, score points, and gain recognition.

Jiang Xu describes the AI circle as "a huge national high school Olympiad training class" — in this kind of environment, intelligence is seen as an identity, while diligence seems to expose one's shortcomings.

But public expression follows a different set of rules. What you say will be nitpicked by strangers, misinterpreted, and contradicted by people who know nothing about the industry. So Founders have to ask themselves first: can I accept being proven wrong?

Founder expression in the AI era is more like a company's "cognitive log".

Not Content: The Hardest Thing for Founders to Outsource Is Their Judgment

Public expression from Founders is so important — why not just leave it to a professional team?

Topic selection, filming, and editing can all be outsourced, and expression skills can be trained. The thing that is truly hard to outsource is why you hold that particular judgment.

Employees can hardly do a good job building a Founder's IP, because the superior-subordinate relationship is inherent.

Employees can polish the boss's words to sound more elegant, but they cannot keep pressing questions like:

"Why do you think that way?" "Is it possible that you are wrong?" "You didn't say that before."

If they dig too deep with these questions, the boss might lose face, there could be violations of hierarchy, and the inquiries might even touch on the company's core interests.

As a result, many Founder IPs end up as something strange: the content team works very hard, the account looks decent, and the performance metrics are acceptable, but if you go through all the content, all you can find is "the boss is always right".

Jiang Xu uses an analogy: platforms can amplify an already well-established creator many times over, but they cannot create one from scratch.

If public personalities could really be mass-produced on an assembly line, viral hits and internet celebrities would have been churned out in batches long ago.

Content formats can be copied, and viral formulas can be learned. But a person's experiences, perspectives, the judgments in their mind, and the willingness to take responsibility for those judgments cannot be produced on an assembly line.

Founder IP is hard to build because it involves so many intertwined factors: the boss's self-esteem, the company's interests, the team's capabilities, and the platform's rules are all packed into a single account. On the surface it looks like a content project, but in reality it is more like the personalized interface of the company.

Content can be produced, but judgment cannot. Therefore, what Founders truly need to deliver to the market is not an account, not fans, not a persona, not perfection, not content — it is their judgment.

But why exactly now?

More than 70 years ago, economist Hayek wrote in *The Use of Knowledge in Society* about an issue that seemed completely unrelated to Founder IP: in a complex society, truly useful knowledge never exists fully in a single brain, but is scattered across countless individuals, across different timelines and scenarios.

Today, AI is continuously learning those scattered, valuable pieces of knowledge, and you can ask it any question and it will quickly give you a seemingly plausible answer.

But when it comes to making real decisions — choosing A over B, when to change your mind, who takes responsibility when things go wrong — these still have to be done by humans.

The better AI gets at answering standard questions, the more valuable human judgment becomes. The market wants to see how you, as an individual, make decisions.

This article is from the WeChat public account "UnDefined", written by Cheng Kele, and published with authorization from 36Kr.