The AI race took a sharp turn on Monday when a Chinese chatbot shook the industry, wiping $1 trillion (approximately Rwf 1.39 quadrillion) off U.S. tech stocks. DeepSeek, a rising Chinese AI firm, is now a serious challenger to consumer favorites like OpenAI’s ChatGPT and Google’s Gemini, fueling concerns that the West could be losing its edge. Hours after DeepSeek’s performance gained global attention, OpenAI rushed out a new ChatGPT version for government agencies. ALSO READ: AI: Who will protect us when we can’t trust what we see or hear? As DeepSeek’s influence grows, we take a closer look at the visionary behind the company. Who’s behind DeepSeek? DeepSeek was founded in Hangzhou by hedge fund manager Liang Wenfeng, who used the proceeds from his financial ventures to fuel the company’s rise. However, some skeptics question the extent of the Chinese government's involvement in its rapid growth. Now serving as DeepSeek’s CEO, Wenfeng is said to have been driven more by ambition than profit. He aims to earn recognition in a US-dominated tech landscape, and his company’s disruptive entry into the AI sector suggests he has done just that. ALSO READ: Where does Rwanda stand in integrating AI, technology? At 40, Wenfeng has already made history. DeepSeek’s chatbot debuted in the United States on the same day as Trump’s inauguration, quickly becoming the country’s most downloaded free app. Its rapid adoption has sent rival AI firms' stock values tumbling and triggered turbulence on Wall Street and in Silicon Valley. Wenfeng’s challenge to US dominance has also earned him national hero status in China. Last week, he met with Prime Minister Li Qiang as the sole AI executive at a gathering of China’s top entrepreneurs. During the meeting, officials reaffirmed Beijing’s commitment to advancing key AI technologies—a signal that DeepSeek’s rise may only be the beginning. A self-proclaimed math ‘nerd’ with a vision for human-level AI, Liang shared his ambitions early in his career but was met with skepticism. “When we first met him, he was this very nerdy guy with a terrible hairstyle talking about building a 10,000-chip cluster to train his own models. We didn’t take him seriously,” a former business partner told the Financial Times. “He couldn’t articulate his vision beyond saying: ‘I want to build this, and it will be a game-changer.’” Born to two elementary school teachers in the port city of Zhanjiang, Liang excelled academically, teaching himself calculus in middle school. Former teachers recall him as a straight-A student who mastered subjects with minimal effort. By his teenage years, he was studying college-level mathematics, eventually earning a degree in information technology from Zhejiang University, one of China’s top institutions. Despite Hangzhou’s reputation as a tech hub, Liang initially pursued a career in finance. He founded Jacobi, an investment firm named after German mathematician Carl Jacobi, before launching High-Flyer in 2015—a hedge fund that used AI-driven algorithms to predict stock market movements. By 2023, Liang pivoted fully to artificial intelligence, founding DeepSeek and recruiting top graduates from China’s elite universities. His hiring philosophy? “Passion and curiosity.” Unlike many CEOs, Liang is deeply involved in daily operations, often working late into the night and even sleeping in the office, according to The Wall Street Journal. DeepSeek’s rapid rise has rattled US tech giants, not only for its impressive performance but also for its relatively small budget. One of Liang’s boldest moves was making DeepSeek’s code open-source—a stark contrast to the closed systems of Silicon Valley. “For technologists, having others follow your work gives a great sense of accomplishment,” he told 36Kr last year. “Open source is more of a culture than a commercial strategy. Contributing to it earns us respect.” Why are US tech firms worried about DeepSeek? DeepSeek launched its first AI model in November 2023, followed by DeepSeek-V2 in May 2024 and DeepSeek-V3 in December 2024. The game-changer came on January 20, 2025, with DeepSeek-R1, which topped the Apple Store’s free apps chart within a week. DeepSeek’s AI matches OpenAI’s latest ChatGPT but costs far less to develop. While US firms spend over $100 million on training, DeepSeek reportedly trained V3 using just 2,000 Nvidia H800 chips—costing only $6 million. Unlike other chatbots, DeepSeek’s AI explains its reasoning before responding. Fully funded by hedge fund High-Flyer, DeepSeek is free from investor pressures, allowing a focus on long-term research. Its team, mainly young graduates from top Chinese universities, prioritizes technical ability over traditional experience. DeepSeek’s emergence could reshape how AI investments are structured, proving that cutting-edge models can be built without massive budgets. Silicon Valley investor Marc Andreessen called DeepSeek-R1 AI’s “Sputnik moment,” referencing the Soviet Union’s launch of the first artificial satellite, which triggered the Cold War space race. “DeepSeek R1 is one of the most amazing and impressive breakthroughs I’ve ever seen—and as open source, a profound gift to the world,” Andreessen wrote in a post. Deepseek R1 is one of the most amazing and impressive breakthroughs I’ve ever seen — and as open source, a profound gift to the world. — Marc Andreessen (@pmarca) January 24, 2025 With DeepSeek defying expectations and challenging Washington’s export controls on advanced chips, US giants like OpenAI, Meta, Google, Apple, and Microsoft now have a serious new competitor on the horizon. Despite the vast resources US firms have funneled into AI, DeepSeek has proven their breakthroughs can be matched. Prof. Neil Lawrence, DeepMind Professor of Machine Learning at the University of Cambridge, believes this is only the beginning. “This is just the tip of the iceberg in terms of the innovation we can expect,” he said. He compared DeepSeek’s impact to James Watt’s role in revolutionizing the steam engine—not by inventing it, but by refining it into something far more powerful.
