Kai-Fu Lee, chairman and chief executive officer of Sinovation Ventures, speaks during the HICOOL Global Entrepreneur Summit on September 11, 2021 in Beijing, China.
China News Service | China News Service | Getty Images
BEIJING — Chinese artificial intelligence models may be at least half a year behind those developed in the U.S., but Chinese AI apps will likely take off much faster, said Kai-Fu Lee, former head of Google China.
He was referring to large language models, which are trained on massive amounts of data that can process and produce text, images and videos.
The top Chinese companies’ LLMs are about six to nine months behind their U.S. counterparts, while less advanced Chinese models may lag the U.S. by about 15 months, Lee said. He was speaking at the AVCJ Private Equity Forum China on Wednesday.
Lee, author of “AI Superpowers: China, Silicon Valley, and the New World Order,“ is a widely followed commentator on AI, and is the founder of startup 01.AI as well as venture capital firm Sinovation Ventures.
“Apps, I would predict, by early next year will proliferate in China much faster than in the U.S.,” Lee said, noting that the cost of training a good AI model has fallen significantly.
“It’s inevitable that China will [build] the best AI apps in the world,” he said. “But it’s not clear whether it will be built by big companies or small companies.”
Lee, whose startup is focused on search apps right how, said it may take five to eight years to take generative AI consumer applications to the next level — a single “super app” that can perform multiple tasks.
The industry will likely need completely new devices versus existing smartphones, he said, adding “the right device ought to be always on, always listening.”
Major Chinese companies such as Alibaba and Tencent have released their AI models and business products. These companies and investors have also backed several AI startups.
Beijing-based ShengShu Technology, backed by Alibaba-affiliate Ant Group, announced Wednesday that its text-to-video model Vidu has introduced a new feature for improving how a main element or character in AI-generated clips can be portrayed consistently, without distortion. That can enable advertisers to create promotional videos for their products.
Vidu was released earlier this year and its basic tools are open to the public, with more advanced capabilities available via subscription. Co-founder and CEO Jiayu Tang told reporters Wednesday that several companies were interested in buying ShengShu’s services, and were not just exploring the tech.
Credit: Source link
