Kimi goes public; Yang Zhilin may become the youngest hundred-billionaire in history

The legendary story of DeepSeek’s founder Liang Wenfeng—top scorer in the national college entrance exam—leaves us awestruck

Don’t rush—amid the tide of the AI era, heroes across the land are like shoals of fish crossing the river

After Kimi’s K3 model was released, it went viral across the entire internet. Within 48 hours of launch, it came close to the limits of compute power, and individual memberships were paused—recreating the same level of hype DeepSeek had back then

Kimi’s current valuation is 200 billion yuan. As K3 goes viral both at home and abroad, there is still room for the valuation to move upward

This June, Kimi’s annualized revenue exceeded $300 million. After K3 is rolled out, revenue is expected to grow by several times. The company could list in Hong Kong in as little as half a year

Based on the current growth rate, if Kimi successfully lists, taking reference from AI market caps in Hong Kong stocks, Kimi’s founder Yang Zhilin is expected to become the youngest “trillion-yuan” billionaire in history

And all of this traces back to the time he refused Apple

Tsinghua “genius”

Most people first came to know Yang Zhilin because of Kimi

But before Kimi exploded, he had already spent many years accumulating in the AI field

Yang Zhilin was born in 1992 in Shantou, Guangdong. He studied for an undergraduate degree at Tsinghua University, and afterward went to Carnegie Mellon University to pursue a PhD

During his PhD, Yang Zhilin took part in research projects such as Transformer-XL and XLNet

These names may be unfamiliar to ordinary users, but they carry some influence in the field of AI research

Especially Transformer-XL, which solved issues that traditional models face when handling long texts. And a few years later, long-form text also became one of Kimi’s earliest tags

However, Yang Zhilin did not choose to keep staying in academia. During his PhD, he already started trying entrepreneurship

In 2016, he helped found Recurrent AI, hoping to use AI to improve efficiency for enterprises

After that, he also joined top research teams such as Google Brain

If he had followed this route, he could have stayed in the United States—entering companies like Apple, Google, and Meta and becoming a top AI research scientist

But in the end, he chose another path: returning to China to start a business

Start a business back in China

In 2023, ChatGPT burst onto the scene, and global AI competition entered a period of intense heat

While giants such as Baidu, Alibaba, and Tencent all jumped in, Yang Zhilin chose to found a startup: Moonshot AI

This name comes from Pink Floyd’s classic album The Dark Side of the Moon, representing the exploration of unknown territories

Compared with many companies at the time that pursued the quick release of a chatbot, Yang Zhilin chose a harder route: starting from the underlying model

In October 2023, Moonshot AI released Kimi

At that time, most large models in the market were still limited to basic scenarios like chatting and writing. Kimi’s earliest differentiating capability was long-text processing

It supports context on the order of 200,000 words, allowing users to directly upload papers, research reports, and long-form materials so the AI can read and analyze them

This capability quickly struck a chord with a large number of knowledge workers

In March 2024, Kimi saw its first major surge. Large numbers of users flooded in, and the servers even, at one point, experienced pressure

From an internal product within the AI circle to a domestically developed large model that ordinary users came to recognize—Kimi only took a few months

Capital also began reexamining the company. Less than a year after its founding, Moonshot AI completed multiple rounds of financing, and its valuation rose rapidly

In the years that followed, the company continued to receive backing from capital. After the latest round of financing in 2026, its post-investment valuation had already exceeded $20 billion, cumulative financing exceeded $3.9 billion, and it became one of the companies with the largest funding scale among domestic large-model startups

If you don’t start a business, you’ll regret it for the rest of your life

As Kimi K3 became a hit, it was also exposed that Kimi’s founder Yang Zhilin’s PhD supervisor at Carnegie Mellon University—Russ—had spoken

He made a statement on social media saying he was honored to have once taught Yang Zhilin, and then a wave of very American-style patriotic critics came out to question him

They said he was training talent for a rival and weakening the United States’ AI strength

He also revealed some details: if Yang Zhilin had been willing to stay in the United States, there were actually many opportunities

Because he had previously been in charge of Apple’s AI laboratory, through that connection, Apple CEO Tim Cook had someone ask Yang Zhilin whether he was interested in joining Apple

But Yang Zhilin’s response was: if he didn’t start an independent business once, he would regret it for the rest of his life

With Yang Zhilin’s credentials at the time, if he stayed in the United States, he would most likely have entered companies like Apple, Google, and Meta, becoming a top-tier scientist

He could earn a high salary, publish papers, and train models—continuously climbing upward within a mature research system. But it would be hard for him to build a model like Kimi

He also said in early interviews that at big companies, scientists can decide how to implement a certain technology, but it is difficult to decide how much compute, funding, and talent a company will commit to researching that technology

Product and business are split across different departments. No matter how strong an individual’s ability is, they are only one part of a massive machine

Another important reason is that if he had stayed in the United States to start a business, he might not necessarily have been able to raise the funding that Moonshot AI later obtained

A Chinese researcher who had just graduated and lacked entrepreneurship networks on the ground in the United States would find it hard for capital to quickly stake huge amounts of money and compute resources on them

There is more money in the United States, but the people who are qualified to get it are not as many as people imagine. After returning to China, the situation turned out differently

Yang Zhilin has both a CMU background and top AI research experience, and he is also familiar with China’s talent, capital, and market

Here, he is not just one scientist in a big-company system, but one of the few people who can connect technology, financing, compute, teams, and product all at once

If Yang Zhilin had stayed in the United States, the world might have gained another AI scientist at a Silicon Valley big tech company

It was only after he chose to return to China that Moonshot AI and Kimi came to be