Baidu CEO Robin Li emphasized on Tuesday that investing in data centers and cloud infrastructure remains essential, even as Chinese AI startup DeepSeek presents new competition. Speaking at the World Government Summit in Dubai, Li stressed that more advanced AI models demand greater computing power—often referred to as "compute"—to function effectively. His remarks come amid growing interest in DeepSeek, which has developed language models capable of delivering performance comparable to OpenAI’s GPT while using significantly less computing power. This has sparked discussions about whether large-scale AI infrastructure is truly necessary.
As a major force in China’s AI sector, Baidu quickly introduced its own AI products following OpenAI’s ChatGPT launch in late 2022. However, its flagship language model, Ernie, has struggled to gain widespread adoption despite claims that it rivals GPT-4 in capability. Li, who had previously asserted that China was unlikely to produce an OpenAI-equivalent company, acknowledged at the summit that AI innovation is inherently unpredictable—citing DeepSeek’s rapid ascent as an example.
Marking a shift in his stance, Li also recognized the potential of open-source AI models in accelerating adoption. While he had previously favored closed-source development, he now sees open access as a way to encourage broader experimentation and faster technological advancement. This represents a notable change in Baidu’s approach to AI’s evolving landscape.
