There are founders who aren’t American tech billionaires like Elon Musk, Sam Altman, or Dario Amodei who are making waves in the AI industry.
Yang Zhilin is the 34-year-old researcher and entrepreneur behind the Chinese company Moonshot AI and its new indiscriminate weight Kimi K3 AI model, which captured the attention of Silicon Valley this week.
The Kimi K3 is the latest trending open weight model from China. Its coding and agent capabilities are comparable to leading US models offered by research institutes such as OpenAI and Anthropic, and at a much lower cost, marking a new milestone for China in the AI race.
Who is Yang Zhilin?
Yang was born in 1992 in Shantou, a city in China’s Guangdong province. He attended Tsinghua University and later received his Ph.D. He studied under prominent AI researchers Ruslan Sarakitnov and William Cohen at Carnegie Mellon University in Pennsylvania.
“He’s really great,” Saravtdinov said of Yan in an X post on Friday. Sarakhutdinov said in another post that Yang was heavily hired by big tech companies after graduation, but was focused on starting his own company.
While attending Carnegie Mellon University, Yang interned at Google Brain and Meta. He has also co-authored several research papers on topics ranging from limitations on how language models handle context to prompt tuning advice.
Before launching Moonshot AI, Mr. Yang returned to China and contributed to several major AI projects, including Huawei’s PanGu model and Beijing Academy of Artificial Intelligence’s large-scale multimodal AI model, Wu Dao. He is also the co-founder of Recurrent AI, a startup that uses artificial intelligence to analyze sales conversations and help companies improve performance.
One of Yang’s principles when building AI is to focus on scale.
“If you can solve it with scale, why not solve it with a new algorithm? The value of a new algorithm is that it allows you to scale better,” he told journalist Xiaojun Zhang in an interview posted on LinkedIn.
He helped launch Moonshot AI in early 2023. The company’s team had developed multiple AI technologies, including Transformer-XL, RoPE, Group Normalization, ShuffleNet, MuonClip, and Mooncake.
Moonshot AI first gained fame by giving the Kimi K2 underlying model an unusually large context window (1 trillion parameters) for processing long documents. Moonshot also has an AI assistant product, Kim, and has since expanded into coding, research and autonomous AI agents, attracting backing from Alibaba, Tencent and other big Chinese investors.
Kimi K3 propelled both Moonshot AI and its founder onto the global stage.
Vercel CEO Guillermo Rauch said in a post on X that Kimi K3 is “the first time an open model outperforms all proprietary models in this comprehensive web engineering benchmark.” Wharton professor Ethan Mollick, meanwhile, called it “the closest we’ve ever been to a frontier.”
“The ultimate AGI company will dwarf today’s giant companies. It will double or triple in size. It won’t necessarily be OpenAI, but such a company will exist,” Yang said in an interview with Zhang.
Why did Yang Zilin leave the United States?
Some key technology insiders lamented that Yang did not stay in the United States to work at an American AI lab after KimiK3 was released this week. Legendary billionaire investor Vinod Khosla has criticized the Trump administration’s restrictive immigration policies.
“The bigger problem is that we are keeping talent away from other countries because of our immigration policies against talent,” he said in response to news of KimiK3’s success.
The Trump administration has moved to tighten entry restrictions, including on student visas. Last year, the government introduced a $100,000 fee for employers who sponsor new H-1B applications for foreign workers. The ruling was later reversed by a federal judge, and the case is still being litigated.
And in May, a U.S. Citizenship and Immigration Services memo suggested that people who previously could apply for green cards from within the United States may now have to leave the country while their cases are processed.
And this week, the administration introduced new rules that put an expiration date on how long people with student visas can initially stay in the United States.
However, Saraftdinov said that while U.S. immigration can be scary and “uncertain,” Yang always intended to return to China and launch his startup.
“I remember him telling me that if I didn’t at least try to start my own company, I would regret it for the rest of my life. I respect that, and he was right,” Sarakhutdinov wrote about X.