The Women Running the Rest of AI Companies

OpenAI has Fidji Simo. Anthropic has Daniela Amodei. One sits next to Sam Altman as the only other CEO. The other is Dario Amodei’s younger sister and Anthropic’s president; everyone except one chief of staff reports to her, and she reports to the board. That is the pattern PANews focuses on in this market analysis: in today’s leading AI companies, women are often handed everything “outside the model.” The article asks what that looks like on the China side, and finds plenty of examples. Zhang Yutong at Kimi, Yun Yeyi at MiniMax and Yi Weishu at Vivix are named as cases in point. Their titles differ — president, CEO, co-founder — but the job shape is similar. Technical founders usually keep the algorithm and R&D side. What remains is a large share of the company: fundraising, people, finance, legal, government relations, PR, commercialization, and, in some cases, product work. In short: the CEO leads the AGI chase; these women make sure the people chasing AGI actually form a company. The article sums it up in one line: “He is somewhere. She is everywhere.” Unlike the classic internet-era No. 2 figure, AI and embodied-intelligence companies often rely on these executives for far more than a standard CFO or COO remit. They oversee almost everything outside the technical core. In some companies, most non-technical teams report to them. Front-line employees may rarely meet the CEO, but they will see these executives again and again in final interviews and business reviews. Externally, investors, clients, media and regulators usually meet them too unless the CEO absolutely has to step in. That work is often described as “everything outside the model,” as if the model were the only real business and the rest were housekeeping. The article pushes back on that idea. Chip procurement and fundraising on the upstream side, consumer growth and enterprise sales on the downstream side — all of it affects whether model training can continue and whether a product survives long enough to reach the next version. Training is central, but success does not happen inside the cluster alone. So these executives are not fully cut off from technical decisions. In some firms, they attend key algorithm and R&D meetings with the CEO and help decide where resources go, when a product should launch, and how long a technical direction deserves another shot. Because they are the ones interacting with the market over time, they also tend to notice shifts in tone faster. Employees quoted in the piece say these executives track competitor training progress, benchmark focus, fundraising and cash runway, then work with the CEO to set direction and push pace. At times, they can look more like the steady manager in the room than the founder does. The article gives one example from a model company where training had stalled for months. The CEO’s rhythm slowed, and employees kept asking what he was doing. A female co-founder eventually confronted him and pushed him to revisit several possible paths. She kept forcing the issue: if the model does not improve, the next round of financing will not come; without financing, there will be no next attempt. They argued many times before the CEO finally moved again. That kind of relationship is hard to capture in a title. These women are neither subordinates nor mere clean-up crews. Often, they have to believe in the founder’s technical judgment while also pulling him back to the table when he loses it or avoids it. Trust, the article argues, is the rarest asset in this role. It is not built by one financing round. Still, fundraising is often the first visible value they bring. And the money is never just money: the nearly $800 million Alibaba brought to Kimi, the financing that helped MiniMax enter the group of China’s leading foundation-model startups at the end of 2022, and Vivivx’s valuation reaching $1.32 billion all carried strategic weight far beyond the cash itself. But fundraising ability is only the visible part. Some investors enter a company and remain fundraising leads; others widen their scope into operations, growth, organization and strategy, eventually helping define what kind of company it should become. The article says this depends on experience, background and, more importantly, whether they can keep evolving. Zhang Yutong is cited as an example. During her time at GSR Ventures, she spent two years as head of operations at FunPlus. At Kimi, her remit did not stop at capital. It expanded from operations and growth outward. The article says that, today, she sounds “less and less like an investor,” which it frames as a compliment. Trust from founders is built the same way: do one thing well, then take on another. Understand the technical ideal, but also translate it into budget, organization, product and revenue. Stand behind the founder when needed, and stand opposite him when necessary. The flip side is that companies without this kind of role often do not fail because they lack someone for fundraising or PR. They fail because they lack someone who can truly hold the organization together. The author even says that on Xiaohongshu, many of the tech companies most frequently criticized for organizational management seem to lack such a figure. The piece also mentions a few rumors: one company had a strong talent pool but severe organizational issues and constant poaching, prompting the top executive to angrily threaten a rival with a lawsuit. Another founder wanted to change direction but could not pull budget and resources back from subordinates after approving them earlier in the year. By then, he realized he was not really managing the company. In another case, a “star team” raised funding and soon saw members looking for bankers and preparing to raise on their own — something many investors have seen before. A high density of technical talent does not automatically produce an efficient organization. In fact, a group of smart, expensive, highly self-driven people often needs someone to manage power, resources and direction. The article then turns back to gender. It argues that the prevalence of women in this role is first of all a demographic and career-path result. Investment, IR and strategy functions already employ many women, so there is no need to explain it with stereotypes like “women communicate better” or “women are more meticulous.” The author also names several male examples, including Han Zheng, CEO of Sudo Technology, and Xie Xuzhang at Aishi Technology. The real point is not gender, but the organizational shift in tech firms. They are less like traditional companies and more like commercializable labs. In the past, CEOs were expected to handle people, money and execution because technology, organization, supply chains and channels could all decide whether a tech company rose or fell. Today, the biggest victory is first and foremost a model or core-technology victory. Technical founders focusing most of their attention on technology is not necessarily a flaw. It may even be the right path. Seen that way, the mid-race outcome in the competition among China’s “six little dragons” of large language models is also, in part, a victory for scientist-CEOs over technically literate managers. CEOs now need to look more than ever like technical idols. Employees at many model companies say they choose jobs based on one question: which company is more likely to reach AGI, and reach it faster. People chasing that goal are following someone they believe can lead them toward a huge ambition. The article says that may also explain why Sam Altman is so strong at communication and commercialization, yet still draws complicated reactions from parts of Silicon Valley’s technical community. But the more technical a company is, the more it needs someone to absorb everything outside the technology. That is why some founders are even willing to give up the CEO title and look for their own “Zhang Yutong.” In 2023, one model-company founder publicly posted a call for a CEO. In casual conversation, several investors also said they had been invited by portfolio companies to take on the CEO role. The article is clear that this does not mean business has been уступed to technology. It means scientist-CEOs know that as long as companies are made of carbon-based life, there will always be human politics, loyalties and rivalries — and technical superiority will not make them disappear.