By Charles Mkoka
The Chinese AI startup DeepSeek made a big splash earlier this year, shaking up stock markets and sparking debates about the competition between the U.S. and China in AI development, according to Africa News.
In the past two weeks, DeepSeek's Chinese competitors have launched over 10 major product updates, some of which have raised concerns about their impact on young users.
DeepSeek's AI assistant became the most downloaded free app on Apple's iPhone store in January, drawing attention as a competitor to ChatGPT.
Many argue that the Chinese startup has caught up with American companies leading the AI field, all at a much lower cost.
While it’s still unclear if DeepSeek’s releases are on par with or better than those from Western companies, they are creating more competition for U.S. businesses.
DeepSeek was founded in 2023 in Hangzhou, China, and released its first AI language model later that year.
The company has said that its recent models were built with Nvidia’s lower-performing H800 chips, which are not banned in China.
This suggests that the most expensive hardware isn’t always needed for cutting-edge AI research, especially with the ban on California-made A100 chips.
DeepSeek started gaining more attention in the AI industry last month when it released a new model that it claimed was as good as OpenAI’s models.
After DeepSeek’s success with open-source technology, OpenAI has hinted that it might also release some of its technology for free, even as it considers charging more for its premium products.
DeepSeek’s chatbot became more accessible when it was made available on Apple and Google app stores earlier this year.
The turning point came when DeepSeek published a research paper on the same day as President Donald Trump’s inauguration.
The paper described a new AI model, R1, which showed advanced reasoning skills, like solving math problems. It was also much cheaper than a similar model from OpenAI called O1.
Chinese companies have historically outperformed global competitors in industries like electric vehicles and solar panels by producing cheaper products at a larger scale. This trend may now be repeating itself in AI.
Global companies, including Apple and BMW, are already working with Alibaba to integrate AI services into their products in China.
Last week, Apple CEO Tim Cook visited Hangzhou, DeepSeek’s home city, to meet with future AI developers at Zhejiang University. During his visit, he donated $4.1 million.
James Wilton, a tech consultant, said, “If you assume the Chinese LLM players’ goal is to disrupt the market and steal market share, it’s working. They might be willing to absorb the costs now, but it won’t stay free forever,” according to Bloomberg.
Balaji Srinivasan, a tech investor, wrote on X, “China is trying to do to AI what they always do: study, copy, optimize, and then drive out competitors with low prices and massive scale.”
One key difference between DeepSeek and its competitors, like OpenAI, is that DeepSeek’s models are open source, meaning anyone can access and modify them for free. However, the company has not shared the data it used to train its models.
