“Core Tips”
It’s starting to be hard to get money. Will the persona of an AI entrepreneur still work?
作者 | 高澤
編輯 | 邢昀
Recently, a set of “Financing Guidelines” has sparked heated discussion in the AI technology venture capital circle:
GitHub needs to accumulate 10,000 stars, the eyes need to be a little more dazed, it is best to always have a Rubik’s cube in your hand, Tsinghua or Stanford printed on the sweater, the computer is older and there are more stickers, and the whole person looks like “there is no need for financing, only the need for computing power.”
This set of guidelines was initially called “Good Price Style”, and later on platforms such as Xiaohongshu, it was called “Hao Rongfeng” with a more straightforward name. It was once a hit in the venture capital circle, with the ridicule of “Dollars will invest in you if they smell technology”, which almost became an industry-level self-deprecation.
The joke is funny because it really hits an embarrassing point in early AI investment and financing:When technology is difficult to verify quickly, the founder’s personality, resume, and narrative often reach investors’ decision-making desks faster than products and data.
As the time comes to the second half of 2026, financing in popular fields such as embodied intelligence has begun to ebb, exit expectations have tightened, and investors have become more cautious in judging projects. When money starts to be picky about choosing projects, there are fewer and fewer things that can be “blown up” by characters and narratives.
The carnival of jokes and the cooling off of financing may seem contradictory, but in fact they share the same logic: the popularity of integrated wind energy is precisely because it reflects the information asymmetry under the black box of technology; and the more the market returns to rationality, the more the space for this method of play is compressed.
1. Want to “price” out and “integrate” into it
“The eyes should be dull, the team should be small, the clothes should be simple, and the words should be able to say, ‘This round is basically full, but I have left a small share for you.’ It is best to put a folding bed in the office. When investors come, they will inadvertently reveal that they have ‘just woken up and were soaking in their workstations.’ When put together, it points to a core persona: I am not here to ask for money, I am here to do things, and I will accept your money by the way.”
A set of “Financing Behavior Guidelines” has received crazy praise in the AI technology venture capital circle.
The source of this guide is an entrepreneur named Dong Haoran. In the conversation with “Leopard Transformation”, she directly showed the original video transcript with timestamps to prove her identity as the “theory proposer”. She initially called this joke “good price style” and transformed the language of “good marriage style” into the venture capital circle version: Silicon Valley cannot screen people by height and appearance, so they can simply use technical indicators such as “GitHub 10,000 stars” to create an equivalent.
Later, this meme spread on Xiaohongshu and other platforms, and the saying “good integration” was added. From “price” to “integration”, the emphasis is actually on the same thing: how to “integrate” yourself and “price” yourself out. According to her explanation, jokes are used to deconstruct a phenomenon, not to follow it; but newcomers who are new to the venture capital circle often regard it as a golden rule.
She herself is more or less the product of this creative logic: she studied physics, graduated from a physics experimental class at a C9 university, and then did research on quantum algorithms for three years in Singapore. She was originally going to study for a Ph.D., but finally quit midway because she couldn’t stand the boredom and wanted to explore the world more, and she started multiple rounds of entrepreneurship.
A joke may become popular not just because it is funny. Behind it is a paradox in early AI investment and financing: AI projects are often a technical black box. What is the capability of the model, whether the inference cost can be reduced, and whether the data quality is reliable…these core judgment basis are precisely the most difficult to fully adjust and quickly verify.
When information asymmetry is pushed to the extreme, some organizations have no choice but to consciously or unconsciously regard the founder’s resume, external image, and grand narrative as a “simple ruler”: first see if the persona is capable of success, and then see if the story is sexy enough.
An investor’s feelings about this are more mixed. He told “Leopard Transformation” that he would indeed encounter founders who used this style of play, but he did not think that anyone was deliberately tricking whom, but ratherAttribute this to the market’s “missed opportunity” anxiety and “deformity” caused by signaling effects.
“If you don’t package it like this, the founders themselves may not know how to raise money; if you don’t screen people like this, investors won’t know who to choose.”
When everyone is using the same words to compete for scarce attention, those who do not follow up may be eliminated by the market first. There is often no clear dividing line between initiative and compulsion. But he also pointed out a standard that is enough to “veto” an entrepreneurial team with one vote – cobbled together: “You can clearly feel that the people in their team are not familiar with them, or they may have just been put together. The hidden dangers of long-term entrepreneurship are great.”
At the same time, the validity period of this set of signals depends on whether the market in which it is located is still unilaterally upward. Starting from the third quarter of 2026, the turning point of the AI track financing market has become increasingly obvious. Money has been withdrawn from “storytelling” projects and concentrated on targets with orders, profits, and technical barriers, showing the characteristics of total shrinkage and structural differentiation.
Especially on the embodied intelligence track, the cold wave suddenly breaks out and the pace slows down significantly. IT Orange data shows that there were 33 financings in the embodied intelligence track in October, with a total amount of 10.105 billion yuan. Compared with the data in June, there were 55 financings, with a total amount of 30.892 billion yuan, a significant decline.
At the same time, the transfer quotations of old shares in the market are concentrated at 6.5% to 20% off, and the targets even include leading companies. IPOs are also slowing down. Reuters and other media reported that the China Securities Regulatory Commission has slowed down the listing process of some humanoid robot companies through informal window guidance, and strictly checks the authenticity of income, landing scenarios, technological innovation, etc.
Quick judgment, quick bets, and mistakes made even after mistakes are made. These conditions for good integration to be effective are gradually loosening.There is nothing right or wrong in the signal itself, but when a market can no longer rely on the next round to cover the previous round of misjudgments, the signal has to give way to something dumber and harder.
2. Zhongguancun and Silicon Valley tell stories together
In fact, before Haorongfeng emerged from the industry, the media had concentrated on the young geniuses in the AI era, and these were also sought after by the venture capital circle.
The stereotype circulating on the Internet is that domestic venture capital investors love to invest in “little geniuses”, and young geeks in their twenties, who are from prestigious schools, or even dropped out of school are the most popular; while Silicon Valley is more tolerant and is willing to invest in those who are more experienced and sophisticated.
This comparison is very popular on the Chinese Internet, but the reality is much more complicated than the joke.
Dong Haoran travels between Silicon Valley and China all year round. In her opinion, both sides actually invest in “little geniuses.” “Silicon Valley also pursues little geniuses, and the Rubik’s Cube is the special skin action for little geniuses in Silicon Valley.” In her description, the fast-turning customized Rubik’s Cube is a very ornamental “performance spectacle.”
VC investment portfolios are inherently diversifying risks;Some people bet on geniuses, while others bet on mature serial entrepreneurs. This money will not only flow to one age group.
The reason why China seems to “only love little geniuses” isLargely because it attracts the most traffic. A young man with a valuation of over 100 million before graduating with a Ph.D. is naturally more likely to attract attention than a silent middle-aged entrepreneur. Traffic amplifies narratives, and narratives in turn strengthen investment preferences.
Dong Haoran has been in Silicon Valley for nearly a year and has met many young geniuses who have “become famous in one battle.” Some people’s deeds are blown to miraculous proportions after being spread through several layers of communication. When the “little genius character” in the joke is taken seriously, the financing brought by attention may cover up the team’s real shortcomings.
Both Zhongguancun and Silicon Valley need stories, but what is more worthy of further study than personal labels such as “little genius” and “Zhongdeng” is the difference in the “genes” of the two sets of funds.
Dong Haoran summarized his observations in one sentence:“Small money is easy to get in Silicon Valley, but big money is easy to get in China.”
The logic behind it is: US dollar funds prefer projects with a large enough market and fast enough growth. If a small product succeeds in the technical direction, it is possible to get money without even entering an incubator. You can get some support by socializing and eating at events in Silicon Valley. Domestic government guidance funds pay more attention to whether the project can be implemented, drive employment, form an industrial chain, and leave assets to the local area. This kind of “big money” is destined to flow to asset-heavy tracks such as embodied intelligence and visible factories and exhibition halls.
On this point, the above-mentioned investor also has the same experience. He does not care about the age of the entrepreneur. He believes that “little geniuses and mid-level entrepreneurs will have the opportunity to get money.” What really matters is whether the progress is fast enough.
He even mentioned an intriguing phenomenon: the Metaverse has only been in the past few years, and there are already many stories about companies that transformed and told world models, and got money again. Narratives will migrate, but capital’s instinct to pursue certainty will not change.
The two narrative systems also gave birth to a whole set of discourse techniques. From “product” to “agent” and then to “OS” and “data flywheel”, it has almost become an unavoidable “triple jump” in AI financing. The advantage of this upgrade narrative is that it is grand enough, but the disadvantage is that it is vague enough. It can accommodate almost any business that has not yet been run through.
The above-mentioned investor said that he is not opposed to talking about a larger vision, but he is wary of another thing:Use grand narratives to cover up real business.
He cited a common phenomenon in the industry:Many people only dare to report ARR (annual recurring revenue) and rarely mention real profits.
“The ARR of a shell product worth tens of millions of dollars may be the best medical beauty treatment for entrepreneurs, but we still have to come back to the question of whether there can be real profits.” He made his words more clear: inflating ARR, anyway, “you are not dealing with the tax department.” If you go on public accounts and podcasts, you can speak less responsibly. This level of moisture is constantly raising the valuation level of the primary market.
In addition to inflating ARR, there is also a more hidden packaging, which is to describe intended orders as existing performance. Framework orders are not equal to actual revenue, let alone delivered contracts. “Orders” of several hundred million may just be a number that can be talked about to the outside world, but never received. The commonality of this type of caliber is to let paper data run ahead of real business, support valuations in the short term, but have to use real money to cash out in the long term.
In the secondary market, this “moisture” is being squeezed out at an accelerated pace. Under the background of “IPO suspension”, the high valuation of the primary market is losing the support of the exit channel.
3. Who can get the next round?
Entering the second half of 2026, financing in some tracks has begun to cool down, and the bubbles that have accumulated previously are being cleared. When hot money exits the market and funds become more cautious in judging prices and businesses, a practical question emerges: Who can get the next round?
To put it more bluntly, tickets for the next round,Changing from “story” to “business”。
Investors’ due diligence radius is getting longer. What they want to look at is no longer just the growth slope, but also harder data such as cash flow, repurchase, and gross profit.
For the above-mentioned investors, the first thing to maintain is valuation discipline. When he encounters a project that “only has PPT and a team, and the valuation exceeds one billion yuan,” he will choose to give up, even if it means missing out.
He is also wary of investments that consider “retirement at intervals” and the transfer of old shares as the main calculation, and is unwilling to use the assumption that the company will be listed in the future to find reasons for the current high price. Just because someone paid for it in the last round is not enough to prove that someone will still be able to pick it up in the next round.
In addition to price, a more basic judgment is whether the business is genuine. He said that what is often done in due diligence is to check whether what the founder said is consistent with what is actually done. “It’s very common for a person to claim to be honest, but to fabricate their income. He also has various reasons and excuses, such as saying he has multiple Stripe accounts.”
Authenticity is important because investors need it to judge whether the team is making progress. If the caliber of revenue and orders keeps changing, the growth shown at each meeting will lose comparability, and it will be difficult for investors to judge how much the company’s capabilities have improved.
The investor admitted that background will make him take a closer look at the founder, but it is “only a condition for judgment, not even completely necessary.” What really makes him willing to bet are those who “think and act at the same time and make obvious progress every time they meet.” They know what they did right and what they did wrong, and they are willing to be honest with business and with their partners. Such performance requires continuous observation and is difficult to accomplish temporarily by preparing for one meeting.
From the entrepreneur’s side, Dong Haoran’s experience provides another perspective. She has started businesses in Singapore, Silicon Valley and Shanghai (Beijing). The first two companies failed to succeed. The voice interactive AI script platform Dramagic she is currently building is the third company she founded. Several entrepreneurial experiences have also made her pay more attention to long-term accumulation and letting others continue to understand the process of doing things.
Looking back on this experience, she attributed it to “saving seven dragon balls” and “asymmetric bets”: a person may not have a single dragon ball when he starts, so he keeps doing it and saving. When he saves five or six dragon balls, the last one or two will come to you – the cost of failure is limited, and the benefits of success can be unlimited. Even if the winning rate is not high, it is worth betting repeatedly.
Related Reading
- Nvidia live updates: Jensen Huang ousted, CFO Colette Kress to serve as interim CEO2026-11-13
- Are mangoes not afraid of red fruits?2026-11-12
- Original: After Huang Renxun, who is most likely to take over Nvidia?2026-11-12
- Samsung’s roll-screen mobile phone is exposed, creating a new form ahead of Huawei and Apple, and folding screens are expected to be eliminated2026-11-12
- Musk’s favorite car is going to be discontinued.2026-11-12
