What is happening on the ground with generative artificial intelligence (AI) in China?
On the one hand, China is still playing catch-up with the United States in terms of large models, such as US-based Open AI’s ChatGPT, the chatbot that broke records for rapid adoption and has ubiquitously captured the imaginations of global executives.
On the other hand, as China is homing in on generative AI, we see a lot of potential for vertical application within industries and other business settings.
According to our research, the current performance of China’s top large language models (LLMs) is roughly equivalent to ChatGPT 3.5 but falls short of the latest offerings from ChatGPT 4.
China is also behind the US in terms of text-to-image and text-to-video large models, whereas the US has been dominating in terms of image (namely OpenAI’s Dall-E 3 as well as Midjourney) and video (OpenAI’s Sora). Turning to multi-modal models (with inputs and outputs in texts, images, and videos), China faces a large gap to catch up with the U.S. The question is how important that is for the years to come. Chinese entrepreneurs are quick to build applications on top of the existing models.
China’s innovators’ have demonstrated a long-term preference for business applications over foundational technologies, although China’s central government is accelerating a shift towards foundational technologies.
China is currently more focused on creating useful applications that are based on existing large models compared with the West. There are a few reasons for this. China’s innovators have demonstrated a long-term preference for business applications over foundational technologies, although China’s central government is accelerating a shift towards foundational technologies. Chinese technologies in sectors such as new energy, new materials, and pharmaceuticals are leading globally. But in the AI space, China has a comparative disadvantage in terms of its access to high-quality data and access to high-performing chip sets from global industry leader Nvidia. (Remember: the US restricted China’s access to its leading AI chip sets last year.)
“From 2014 to 2023, China-based inventors filed patents for more than 38,000 GenAI-related inventions – six times more than the second-place US with 6,276.”
Already, China has standouts in GenAI business applications, ranging from autonomous driving (thanks to carmakers BYD, NIO, Xpeng, and others) to AI chatbots in retail and e-commerce (thanks to giants such as Alibaba and Bytedance).
Consider this: Over the past decade, according to recently released United Nations data, China has led the world in terms of AI patents filed. From 2014 to 2023, China-based inventors filed patents for more than 38,000 GenAI-related inventions – six times more than the second-place US with 6,276.
This follows a pattern, as China continues to excel in finding applications for new technologies. China’s superpower is its ability to build and commercialize applications at great speed. GenAI applications are no exception. For decades, Chinese companies and entrepreneurs have been working to become masters of applications by recombining new and existing technologies and business models.
In 2023, the total private investment in the generative AI field in China was about $650 million, compared with $22.46bn in the US, according to the Stanford 2024 AI Index Report.
The main players in the Chinese GenAI space can be divided into four groups:
In the startup space, we are watching Moonshot AI, Minimax AI, ZhiPu AI, Baichuan AI, 01.AI, and Modelbest. These six startups have benefitted from venture capital inflows in the past few years, with deal sizes over U.S. $200m. The privately held Moonshot AI is now said to be worth over $3bn.
We see a lot of movement from China’s top technology companies and startups in both open-source and closed-source large models. Although this is an area where China is playing catch-up, it is of strategic importance to spur future growth.
In 2023, the total private investment in the generative AI field in China was about $650 million, compared with $22.46bn in the US, according to the Stanford 2024 AI Index Report. But as China’s central government is prioritizing GenAI solutions to help boost its economy and efficiency amid a shrinking workforce, the West should keep watch. With the central government’s support, local governments are setting ambitious objectives at a provincial or city level. For example, Nanjing city set a 2024-2026 policy goal:
The larger picture is clear: Chinese GenAI is worth following.
AI x 9: This article appears in a nine-part summer series examining how AI impacts leadership and business, produced in collaboration with Expansión.
Companies are rushing to embrace the game-changing opportunities to improve operations and scale that AI offers; yet many are unaware of the pitfalls. Test your knowledge of the risks here – and...
In an era defined by advanced AI and behavioral data, Apple’s 'Evil Steve' test exemplifies a proactive approach to ethical decision-making and responsible data governance....
In a world of deepfakes, hallucinations, and bias, we offer practical guidance on how you can trust your AI systems....
The race to harness artificial intelligence is reshaping the financial services landscape, with industry leaders demonstrating how strategic AI implementation drives competitive advantage. IMD's AI Maturity Index reveals the winning formula. ...
Explore first person business intelligence from top minds curated for a global executive audience