Former Google CEO Eric Schmidt has issued a warning that China may outpace Western countries in open-source artificial intelligence (AI) if the West does not prioritize it. Speaking at the AI Action Summit in Paris, Schmidt highlighted that most leading American large language models (LLMs), such as Google's Gemini, OpenAI's GPT-4, and Anthropic's Claude, remain closed-source.
Schmidt expressed concern that if the West does not act, China could become the leader in open-source AI while other parts of the world remain dependent on proprietary systems. This could pose challenges for Western universities and researchers, who might find it financially burdensome to access proprietary AI models, potentially stifling scientific discovery. While Meta's Llama is an exception, Schmidt believes more effort is needed to maintain technological leadership.
His warning comes on the heels of significant advancements by Chinese startup DeepSeek, which recently launched a powerful open large language model, R1. Last month, DeepSeek's introduction caused ripples in the tech market, leading to fluctuations in AI stocks, including a notable 17% drop in Nvidia's (NVDA, Financial) stock price.
Notably, during the AI Action Summit, the U.S. and the UK diverged from over 60 countries, including China, India, and Germany, by not signing a global AI safety declaration. OpenAI CEO Sam Altman recently acknowledged a shift in perspective towards open-source, suggesting the need for a new strategy. Schmidt, however, advocates for a balanced approach, combining both open-source and closed-source models.
