At the World Government Summit in Dubai on Tuesday, Baidu CEO Robin Li emphasized the continued importance of investing in data centers and cloud infrastructure despite DeepSeek challenging the cost efficiency of large AI models.
Li stressed the necessity of investing in cloud infrastructure, stating, "The investment in cloud infrastructure is still very much required. In order to come up with models that are smarter than everyone else, you have to use more compute."
Computing refers to the hardware resources essential for AI models to train on data, process information, and make predictions.
Li's remarks coincide with DeepSeek, a Chinese AI startup, developing language models that rival top systems like OpenAI's GPT while using significantly less computing power, sparking debate about the need for extensive AI infrastructure spending.
Despite Baidu being an early adopter of AI products post the release of OpenAI's ChatGPT in late 2022, their large language model Ernie, said to match GPT-4's capabilities, has had limited public uptake.
Previously known for bold statements about China's AI sector, including doubting the emergence of another OpenAI-like entity in China and advocating closed-source models as the sole AI development path, Li acknowledged DeepSeek's sudden rise as a symbol of innovation's unpredictability.
Reflecting on the need for innovation due to compute constraints driving Chinese companies to seek cost reductions, Li's stance seemed to evolve towards open-source development, recognizing its potential in accelerating AI adoption.
"If you open things up, a lot of people will be curious enough to try it. This will help spread the technology much faster," Li concluded.