At the age of 24, she raised 400 million yuan in financing.
There are more and more post - 2000 faces in the venture capital circle.
According to investment industry sources, Axiom Math has officially completed its first - round financing of $64 million (approximately RMB 460 million). The financing was led by B Capital, with participation from institutions such as Greycroft, Madrona, and Menlo Ventures. After the investment, the company's valuation reached $300 million (approximately RMB 2 billion).
The founder of Axiom is Carina Hong, a post - 2000 entrepreneur. Born and raised in Guangzhou, she attended the prestigious South China Normal University Affiliated High School and won multiple medals in Olympic mathematics competitions. Later, she was admitted to the Massachusetts Institute of Technology, graduated with a master's degree from the University of Oxford, and then went to Stanford University to pursue a doctoral degree. Dramatically, the birth of Axiom came from a conversation in a coffee shop.
(Photo source: Massachusetts Institute of Technology)
Quietly, post - 2000 founders are making their way onto the AI stage in large numbers.
10 people complete the first - round financing
Valued at $2 billion
Who is Axiom?
The official website shows that this emerging AI company positions itself as "starting with an AI mathematician to build a self - improving super - intelligent reasoning system." In other words, it is a model that can solve complex mathematical problems, generate detailed reasoning for different steps taken to obtain answers, and conduct verification at the same time.
Specifically, Axiom aims to convert English mathematical content from textbooks, papers, and journals into procedural knowledge so that AI can solve mathematical problems and verify the solutions. In the future, the research scenarios of this model are expected to expand to fields such as financial modeling, chip architecture, and even quantitative trading. In the vision of founder Carina Hong, this model may propose new mathematical conjectures and generate brand - new knowledge.
Why target the AI mathematics track? There is an episode. Earlier this year, ChatGPT o3 was reported to have cheated in a math test. Carina Hong pointed out on social media that the reason why OpenAI's large - language model performed well in the math test might be that the company had access to these questions during model training.
She gave an example. In the American Invitational Mathematics Examination, some large - language models could achieve an accuracy rate of 96%. However, when asked to show the proof process, the score dropped sharply to 5%. "Why is there such a difference? Because the training method is flawed." The combinatorial mathematics problems that current large - language AI models cannot perfectly solve are precisely Carina Hong's research specialty.
In Axiom's view, AI has brought about a historical turning point, where discoveries in the field of mathematics can develop at an exponential rate. Therefore, we are now on the threshold of a mathematical renaissance. This renaissance is built on three pillars: AI, programming languages, and mathematics. Together, they drive a series of discoveries, and each breakthrough provides strong impetus for the next innovation.
Now that the first - round financing is officially announced, Axiom's core team has also made its debut for the first time. Although there are currently only 10 full - time employees, there are many AI experts from Meta among them.
For example, the current CTO, Shubho Sengupta, once led the Meta FAIR team and developed OpenGo and CrypTen. He also participated in the research of Google's distributed training system and was one of the earliest experts to develop CUDA technology. François Charton was in charge of the research on large - language models in mathematics and theoretical physics at Meta. He began exploring how to apply Transformer technology to complex mathematical problems in 2019. Hugh Leather, a former AI research scientist at Meta, developed some of the first large - language models for compiler and GPU code generation.
As the leading investor in this round, a partner at B Capital said on the official website that solving complex mathematical problems is at the core of many human inventions. Therefore, a tool that can propose new quantitative hypotheses and verify them, providing a reliable basis for decision - making, will undoubtedly be of extremely high value. The Axiom team also has unique advantages to seize this opportunity and turn advanced mathematical reasoning technology into reality.
"Carina Hong is an outstanding mathematician. She not only has remarkable academic achievements but also has the enthusiasm and focus of the best entrepreneurs. She combines visionary leadership with unwavering execution perfectly." Her journey has just begun.
Led by a post - 2000 female academic superstar
From Guangzhou
The story of Axiom is inseparable from Carina Hong, who has had an outstanding academic journey.
Born in Guangzhou, she is 24 years old this year. Her parents are from Chaoshan. Although her parents did not receive higher education, Carina Hong showed extraordinary mathematical talent from a young age. She attended the South China Normal University Affiliated High School, a well - known school in Guangdong. In high school, she was one of only four girls in the provincial math olympiad team and won good rankings in competitions such as the "Hua Luogeng Cup" and the National High School Mathematics League. In 2019, she was admitted to the Massachusetts Institute of Technology to pursue a double major in mathematics and physics.
During her undergraduate studies, Carina Hong served as the president of the mathematics association, took 20 postgraduate and doctoral courses, and published many high - quality academic papers in fields such as the L - functions of modular elliptic curves and K3 surfaces, the moonshine conjecture, and research on the pop - stack sorting algorithm. Before and after graduation, she first won the Schafer Award for Excellence in Mathematics, which is awarded to only one undergraduate female student each year. Later, she won the Morgan Prize, the highest honor for North American undergraduate mathematics students. She is the fifth female to receive this honor.
At the end of 2022, Carina Hong successfully won the Rhodes Scholarship from the University of Oxford, becoming one of only four Chinese recipients. The Rhodes Scholarship is one of the oldest and most prestigious international scholarship programs in the world, known as the "Nobel Prize for undergraduates."
Subsequently, Carina Hong went to the University of Oxford in the UK to study neuroscience and obtained a master's degree. During this period, she also conducted research on AI and machine learning at University College London. At that time, she said, "How will AI interact with scientists in the future? This is the topic I hope to study next."
In August 2024, Carina Hong entered Stanford University to pursue a doctorate in mathematics and a juris doctor. "I've always been a researcher," she said in a previous interview. "Deep down, I want to solve truly difficult technical problems." Soon, while still a doctoral student, she entered the AI field.
She founded Axiom after a conversation in a coffee shop.
One weekend last fall, Carina Hong had a conversation with Shubho Sengupta, who was then working at Meta, in a coffee shop near Stanford. They talked for several hours, discussing the intersections of their research fields and the possibility of developing AI to solve the world's most challenging mathematical problems.
Shortly after this conversation, Carina Hong dropped out of Stanford and started Axiom. Interestingly, Axiom's meeting rooms are named after great mathematicians, such as Gauss and Ada Lovelace.
When DeepSeek became a global sensation at the beginning of the year, Carina Hong said with emotion, "A small, focused, and maverick team. An excellent group of idealistic partners. They have strong execution ability and are hands - on. Most importantly, they have a belief that combines ideals and missions. This is the story of DeepSeek, and it's also the story I want to write myself."
As Axiom states on its official website, "The future of mathematical discovery starts here."
Joining the AI field
Post - 2000 entrepreneurs are making their collective debut
Looking around, post - 2000 founders are collectively taking the stage in the AI field.
Not long ago, two post - 2000 entrepreneurs from MIT, 22 - year - old Chinese - American Jessica Wu and 23 - year - old Neil Deshmukh, founded Sola Solutions and received financing from well - known Silicon Valley venture capital firms. According to the official website, the financing includes a $3.5 million seed round led by Conviction and a $17.5 million Series A round led by a16z with Conviction participating, totaling $21 million (approximately RMB 150 million).
Such cases are emerging.
In May this year, AI startup Anysphere completed a $900 million financing round (approximately RMB 6.5 billion), with a valuation of $9 billion, becoming one of the hottest AI programming unicorns this year. The company is led by four young faces. In 2022, Michael Truell, who graduated from MIT, and his three classmates decided to start a business in the AI programming field. Their product, Cursor, has changed the way of programming and quickly swept through the Silicon Valley AI circle.
Similarly, in March this year, AI recruitment website Mercor announced the completion of a $100 million Series B financing round, with a valuation of $2 billion. The growth rate is astonishing. The company was founded by three post - 2000 entrepreneurs who dropped out of school. In their sophomore year, they founded Mercor in their dormitory and later decided to drop out of Harvard and Georgetown University to focus on entrepreneurship full - time.
Looking back at the domestic situation, a similar scene is also playing out.
At the beginning of this year, three post - 2000 "geeks" from Tsinghua University, Min Yuheng, Cheng Yi, and Li Yizhe, started a business in robotics, and Lingcifang was born. Now, the company has successfully completed the angel + and angel ++ rounds of financing worth hundreds of millions of yuan, attracting well - known institutions such as Jijie Capital, Tongchuang Weiye, Lihe Kechuang, Shuimu Fund, Mizuho Lihe, and Ralph Venture Capital.
There is also Lingchu Intelligence, which completed angel - round financing led by Hillhouse Capital and BlueRun Ventures. Co - founder Chen Yuanpei is a post - 2000 entrepreneur obsessed with robotics. He was a visiting scholar at Stanford University, studying under professors Karen Liu and Fei - Fei Li. Yang Fengyu, the founder and CEO of the embodied intelligence company UniX AI, was born in 2000. He graduated from the University of Michigan with a major in computer science and returned to China to start a business after obtaining a doctorate from Yale University. His company has also attracted the attention of many VCs.
A post - 2000 army is quietly rising. Some investors have analyzed that in this wave of AI entrepreneurship, post - 2000 founders have the same starting point as post - 90s and post - 80s founders, and they may even have greater advantages. Since many things are too new, previous experience cannot be directly applied, and post - 2000 entrepreneurs have no baggage and have a better understanding of new things.
"We always believe that entrepreneurship often belongs to the young," said Dai Yusen, a managing partner at ZhenFund, recently. "The ignorant are fearless. Many scientific and technological innovation achievements are born out of the fact that people are not afraid despite understanding the difficulties, thus giving rise to innovative ideas."
"The young carry high aspirations and will surely shine one day." Every generation of young entrepreneurs is destined to leave their mark in the long river of history.
This article is from the WeChat public account "Investment World" (ID: pedaily2012). The authors are Liu Bo and Wang Lu. It is republished by 36Kr with permission.