Now, it has been more than two years since the first human-like generative AI called ChatGPT was released. At that time, people were fascinated by its capabilities because they all saw how remarkable and helpful their new e-friend could be. That was true until now. A couple of days ago, China finally published its version of the generative AI model called DeepSeek. By open-sourcing its model and making it available to public use, China’s AI company published a paper in which they measured the model’s performance. Data suggests that the DeepSeek-R1 model outperformed ChatGPT o1 in many important metrics.
This publication of DeepSeek-R1 can mean only one thing. China wants to show the USA that not only do they follow the current trends in IT, more specifically AI, industry but that they are advancing the field and that they have even more powerful tools. The release of this model has brought other benefits to China such as an economical one. Now, since DeepSeek-R1 is open-sourced, many Chinese companies can use this new state-of-the-art model for free instead of paying enormous amounts of money to foreign companies such as OpenAI.
In order to better understand how it is possible for a Chinese startup to develop a model that surpasses the one from the leading tech companies, we would have to start with the company’s founder and CEO, Liang Wenfeng. Liang had been working for one of the largest hedge funds in China for many years before he created DeepSeek. While working for the hedge fund, he was buying tons of GPUs in order to improve computational power required for enormous loads of data analysis. In 2022, he realized that with all the money and available computation power could build cutting-edge models and hopefully develop artificial general intelligence. Today, DeepSeek is one of the only leading AI firms in China that doesn’t rely on funding from tech giants.
DeepSeek’s research team consists primarily of recent PhD graduates from top Chinese universities like Peking and Tsinghua, chosen for their academic excellence and eagerness to prove themselves rather than industry experience. This approach that the company’s CEO fosters evolves into a healthy and collaborative culture, enabling researchers to pursue unconventional projects with sufficient computing resources. Liang believes young researchers excel in high-investment, low-profit research due to their mission-driven mindset. Their education in China and sense of patriotism, especially amid U.S. tech restrictions, fuels their determination to solve critical challenges and elevate China’s global standing.