Chinese AI startup 01.AI, founded by computer scientist Kai-Fu Lee, launched its new AI model, Yi-Lightning. It is touted to surpass OpenAI’s GPT-4o-2024-05-13 and Anthropic’s Claude 3.5 Sonnet on the LMSYS leaderboard. However, the latest ChatGPT-4o still ranks at the top.
Focuses on Enterprise
01.AI is refocusing on enterprise solutions in China, as consumer products have been tough to monetise. Its chatbot, Wanzhi, saw only 100,000 visits in September, compared to over 20 million for rivals Baidu’s ERNIE Bot and Moonshot’s Kimi, per SimilarWeb.
On Yi-Lightning’s launch day, 01.AI introduced a digital human solution for retail and e-commerce. While shifting to enterprise in China, 01.AI remains focused on global consumer apps like PopAI, Monoland, and the AI search app BeaGo.
Yi-Lightning’s strong ranking on the LMSYS leaderboard, surpassing GPT-4o, shows China’s growing strength in AI. With lower costs and faster performance, Yi-Lightning is appealing to developers, especially in markets where affordability is key.
In a previous interview, Lee stressed the need for China to develop its own version of ChatGPT, OpenAI’s 2022 chatbot, which is banned in the country, to boost interest, adoption, and investment.
China’s AI Evolution
Hugging Face CEO Clement Delangue praised Chinese AI progress on X, stating, “Qwen 72B is the king, and Chinese open models are dominating overall.” Alibaba’s Qwen 2-72B now tops Hugging Face’s LLM Leaderboard, outperforming other open-source models.
“China’s advantage is doing whatever it takes to catch up,” said Lee.
Chinese tech companies have also been receiving support from the government in its ongoing AI battle against America. The Chinese government has made AI a national priority, aiming to become the world leader in AI by 2030.
The post Kai-Fu Lee’s AI Startup Claims to have Surpassed GPT-4o appeared first on AIM.