In January of 2025, DeepSeek, a Chinese artificial intelligence (AI) startup, released its latest AI model, DeepSeek R1. The chatbot closely resembles ChatGPT and other related AI chatbots but is significantly cheaper to develop. This breakthrough has dominated global headlines and topped download charts on the Apple App Store.
DeepSeek was founded in December 2023 by Liang Wenfeng. According to BBC, Mr. Liang graduated from Zhejiang University and studied computer science and electronic information engineering. In a rare interview last year, he stated that China’s AI sector “cannot remain a follower forever” of U.S. AI development, highlighting Chinese companies’ growing ambition in the industry.
DeepSeek’s R1 model is open-source, meaning that users are able to access and modify the code to enhance its capabilities. According to Built In, R1 is the latest in a series of AI models released by DeepSeek, but none of its predecessors have garnered as much attention. The company claims that its R1 model is able to excel at both English and Chinese text prompts. However, it is also supposedly good at mathematical calculations and breaking down complex scientific concepts.
However, like other AI models, DeepSeek R1 has its own limitations. Built In reports that the model sometimes mixes languages in its response, especially if the prompt isn’t given in English or Chinese. In addition, DeepSeek struggles with prompt engineering, where users provide specific examples to guide the chatbot’s responses.
DeepSeek’s entrance into the technological world has notably affected the AI industry, leading individuals to reconsider their traditional beliefs about artificial intelligence. Since the release of OpenAI’s ChatGPT model in late 2022, numerous Chinese tech firms and companies have rushed to create their own chatbots that would operate similarly to the widely acclaimed ChatGPT model. However, the results were underwhelming, with large disparities between ChatGPT’s features and the AI models created by Chinese firms. The prevailing belief worldwide was that a significant amount of funding and resources would be required to create a chatbot model comparable to ChatGPT.
However, DeepSeek’s newest AI model stands as an outlier to past failures by various Chinese firms. The company has been especially praised for not only its quality but also for the efficiency of its model despite having significantly lower costs, as stated by Reuters. The rapid yet immense success of the newest AI model has exemplified that high-performing AI models can be developed at much lower costs and less computing power than traditionally believed. According to The Motley Fool, while it cost OpenAI around $20 billion to create the newest model of ChatGPT, DeepSeek’s newly-released V3 model costed only $5.6 million. Despite the colossal disparity in cost, DeepSeek’s performance has proven competitive with OpenAI’s GPT-4o models on multiple occasions.
DeepSeek’s revolutionary model has also triggered great changes in the stock market. Most notably, on January 27, 2025, the stock price of Nvidia, one of the largest listed tech companies in the U.S., plummeted by a shocking 17% in a single day after the release of DeepSeek’s A1 model. Americans began to express potential concerns about Nvidia’s future in the AI industry, fearing that its decade-long dominance in the AI and chip market was gradually slipping away.
The arrival of DeepSeek’s AI model to the tech industry has undoubtedly had significant impacts on the stock market and, more importantly, on public perceptions of the future of the AI industry. DeepSeek’s rise to prominence as a leading tech company in China is expected to continue, potentially ushering in a future where AI becomes more accessible and cost-effective on a global scale.
