Eric Schmidt warns West to focus on open source AI that competes with China
Free updates are notified at any time
Just register AI Myft Digest – Deliver directly to your inbox.
Former Google chief Eric Schmidt warned that Western countries need to focus on building open source artificial intelligence models or risk losing to China in a global race to develop cutting-edge technologies.
The warning comes after Chinese startup DeepSeek shocked the world’s R1 last month, its powerful open-end large language model is more effective with U.S. rivals such as OpenAI, such as Openai.
Schmidt has become a major tech investor and philanthropist, and he says most of the highest LLMs in the U.S. are closed – meaning everyone can’t access freely – including Google’s Gemini, Anthropic’s Claude and OpenAI’s GPT-4, while the exception is Meta’s Llama.
“If we don’t do something about it, China will eventually become an open source leader and the rest of the world will be a closed source,” Schmidt told the Financial Times.
Billionaire says failure to invest in open source technology will prevent scientific discoveries from happening in Western universities, which may not afford expensive closed models.
Schmidt spoke at the AI Action Summit in Paris this week, which also led U.S. Vice Chairman JD Vance Vow to say the United States will remain the main force of the technology.
Openai chief Sam Altman said last month that he was on the “wrong side of history” when it comes to open source models, suggesting that the company needs to develop a new strategy to stay away from expensive closed models.
However, the group has made early progress in developing the technology, resulting in talks with SoftBank’s massively funded startups and a box office valuation of $260 billion. Other big American tech giants, including Google and Amazon, have also invested billions of dollars, believing that this will be the best way to secure a large return on investment.
“I think [Altman] Schmidt said: “It’s a little too fast.
Although the United States dominated the first phase of AI development by building powerful AI models such as GPT-4.
“Europe must be united,” he said. “The application layer is very strong and it will make your Europe more effective.”
On Wednesday, Schmidt announced that he would invest $10 million through Schmidt Sciences to develop a new AI security science program, a nonprofit he established with his wife Wendy.
The program will include 27 projects to engage in basic research on AI security issues. Winners include Turing Award-winning computer scientist Yoshua Bengio, who will develop risk mitigation technologies for AI Systems, and Zico Kolter, a member of the OpenAI board and professor at Carnegie Mellon University, who will explore AI attacks.
Schmidt called on the West to cooperate with the Chinese on AI security as the country will face the same problems around powerful technology. For example, he shares it with the type of information the army does when testing rockets.
“It might be unfavorable to give us information that can be used to make the model safer?” he said.
Other reports by John Thornhill
