Advertisement
Artificial intelligence
TechTech Trends

OpenAI’s GPT-5 draws mixed reviews in China amid heightened AI competition

GPT-5 is not significantly ahead of Chinese AI models, according to Zhang Linfeng, an AI professor at Shanghai Jiao Tong University

Reading Time:2 minutes
Why you can trust SCMP
10
OpenAI says GPT-5 is its smartest, fastest and most useful AI model. Photo: Shutterstock
Ben Jiangin Beijing
OpenAI’s latest flagship artificial intelligence model, GPT-5, has drawn mixed reviews in China, where some critics expressed disappointment over the new system’s lack of breakthroughs.

At its live-streamed launch on Thursday in the US, GPT-5 was touted by OpenAI as its “smartest, fastest, most useful model yet, and a major step towards placing intelligence at the centre of every business”.

The new AI model features improved performance across coding, maths, writing, health and visual perception, among others. OpenAI described it as “a unified system” that features a built-in “thinking” function, with the ability to automatically switch between “standard” and “deep thinking” modes based on factors such as conversation, task types and query complexity.

Advertisement

The new system is “like a PhD-level expert in anything, any area”, OpenAI CEO Sam Altman said at the launch.

In mainland China, where ChatGPT and other OpenAI services are not officially available, AI experts were confident that domestic users would not miss out on anything.
Advertisement
“GPT-5 is not significantly ahead of Chinese models, so it won’t put substantial pressure on Chinese researchers and developers,” said Zhang Linfeng, assistant professor at the School of Artificial Intelligence at Shanghai Jiao Tong University, in a Saturday post on Xinchuang Shanghai, a WeChat public account affiliated with the state-backed newspaper Jiefang Daily.

GPT-5 “doesn’t come with revolutionary breakthroughs; it lacks memorable characteristics”, Zhang said.

Advertisement
Select Voice
Choose your listening speed
Get through articles 2x faster
1.25x
250 WPM
Slow
Average
Fast
1.25x