▲ Moonshot AI Founder Draws Attention
As interest grows in Moonshot AI, the Chinese artificial intelligence (AI) startup that stunned Wall Street and Silicon Valley with the release of its massive open-weight model "Kimi K3," the background of its founder, Yang Zhilin, has also become a hot topic.
It has been highlighted that despite being in a position to easily settle in the United States after earning his doctorate there, he chose to return home and successfully launch his startup in China.
A post on X (formerly Twitter) questioning why a top-tier talent would leave the U.S. has garnered over 7 million views as of July 20.
Yang, a native of Guangdong Province in southern China, graduated from the Department of Computer Science at Tsinghua University before moving to the U.S. to earn his Ph.D. from the School of Computer Science at Carnegie Mellon University (CMU).
Contrary to expectations that he would join a major U.S. tech firm after graduation, his final decision was to head to China.
His goal was to return to his home country and start his own company.
Having already made a name for himself in the U.S. AI industry through stints at Facebook AI Research and Google's deep learning research team, Google Brain, he was reportedly a talent that even Apple sought to recruit.
His voluntary return to China has drawn particular attention, as the papers he published during his doctoral studies are still considered to hold a significant position in large language model (LLM) research.
Analysts suggest that his decision to return and start a business aligns with the Chinese government's active efforts to cultivate and support talent in science and engineering at the national level.
In an interview with local Chinese media this March, Yang said, "China has built the world's best talent system over several decades," adding, "From compulsory education to high school, university, and doctoral programs, we are able to cultivate people with the strongest investigative capabilities in the world, and a system of this kind has already begun to take shape."
His doctoral advisor also refuted public speculation that Yang returned to China due to the complex and difficult immigration policies of the U.S. government, revealing that Apple had indeed offered him a position.
Ruslan Salakhutdinov, his Ph.D. advisor at Carnegie Mellon University, explained on his X account that "Yang Zhilin had plenty of opportunities to stay in the U.S. if he had wanted to," noting that Apple had inquired about his interest in joining, but Yang was determined to return to his homeland.
He added, "An executive (at Apple) even said he could work from the office in Beijing, but Yang was very firm in his resolve to go back and build a startup."
He founded Moonshot AI in Beijing in 2023, and just three years later, on July 17, he released an AI model that shook the world.
Moonshot stated that "Kimi K3" outperforms all competing models in overall performance, with the exception of Anthropic's "Claude 3.5" and OpenAI's "GPT-5.6."
K3 is considered to be of an exceptional scale for an open-weight model, with 2.8 trillion parameters.
In an article on July 19 (local time), the U.S. business news outlet Business Insider introduced him, saying, "There is a founder who caused a sensation in the AI industry this week," and described him as "a 34-year-old researcher and entrepreneur leading the Chinese AI company Moonshot AI, not a billionaire tech entrepreneur from the U.S. like Elon Musk or Sam Altman."
Following the success of "Kimi K3," foreign media reported that Moonshot is pushing for an IPO on the Hong Kong stock exchange within six months.
Bloomberg reported, citing sources, that Moonshot recently officially distributed a resolution to shareholders seeking approval for a Hong Kong listing.
As "Kimi K3" delivered a shock to the market comparable to the "DeepSeek shock"—a generative AI model from a Chinese startup that appeared in January last year—analyses comparing Yang Zhilin to DeepSeek founder Liang Wenfeng have also followed.
Unlike Yang, Liang is a homegrown Chinese talent who did not study abroad in the U.S., but both share the commonality of being strictly researcher-type founders.
They are drawing attention for their path, which differs from so-called "star CEOs" who put themselves at the forefront.
(Photo: X (formerly Twitter) capture, Yonhap News)
※
Copying, redistribution, and unauthorized use in AI training are strictly prohibited.