Zhipu AI COO: Five Years Needed to Master Industry at BEYOND Expo 2025
At BEYOND Expo 2025, Zhang Fan, Chief Operating Officer of Zhipu AI, discussed the company's strategic direction and the broader artificial intelligence landscape. Zhipu AI, a leading Chinese developer of large language models founded in 2019, has recently filed for IPO guidance, positioning itself as a potential pioneer in public listings within its sector. Zhang highlighted the industry's transition from converting computing power into intelligence to transforming that intelligence into scalable productivity. He referenced OpenAI’s five-level AGI framework and acknowledged the positive market impact of competitors like DeepSeek. While noting that model capabilities continue to evolve, particularly in inference performance, Zhang emphasized Zhipu’s core competency in optimizing foundational models and reducing deployment costs. Crucially, he stated that Zhipu will not develop end-use applications internally, arguing that it takes at least five years to truly understand any specific industry. Instead, the company aims to partner with domain experts to combine foundational AI with sector-specific knowledge, thereby generating practical, scalable solutions for enterprise adoption.
Wire timeline
Zhipu AI COO: Five Years Needed to Master Industry at BEYOND Expo 2025
At BEYOND Expo 2025, Zhang Fan, Chief Operating Officer of Zhipu AI, discussed the company's strategic direction and the broader artificial intelligence landscape. Zhipu AI, a leading Chinese developer of large language models founded in 2019, has recently filed for IPO guidance, positioning itself as a potential pioneer in public listings within its sector. Zhang highlighted the industry's transition from converting computing power into intelligence to transforming that intelligence into scalable productivity. He referenced OpenAI’s five-level AGI framework and acknowledged the positive market impact of competitors like DeepSeek. While noting that model capabilities continue to evolve, particularly in inference performance, Zhang emphasized Zhipu’s core competency in optimizing foundational models and reducing deployment costs. Crucially, he stated that Zhipu will not develop end-use applications internally, arguing that it takes at least five years to truly understand any specific industry. Instead, the company aims to partner with domain experts to combine foundational AI with sector-specific knowledge, thereby generating practical, scalable solutions for enterprise adoption.
TechNode