What's Happening?
Chinese AI company Zhipu is making significant strides in the artificial intelligence sector with its GLM 5.2 model, which is closing the gap with leading American AI models on key benchmarks. The model is notable for being free and open-source, which has
facilitated its rapid adoption compared to other models like DeepSeek. CNBC's Deirdre Bosa reports on the implications of this development for enterprises and vertical AI applications. The report includes insights from Box CEO Aaron Levie on model selection, Harvey's Gabe Pereyra on building atop open-source platforms, and Bernstein's Stacey Rasgon on the impact of OpenAI's new Jalapeño chip on the inference-cost race affecting companies like Nvidia and Broadcom.
Why It's Important?
The advancement of Zhipu's GLM 5.2 model represents a significant challenge to American AI dominance, particularly in the context of open-source technology. This development could lead to increased competition in the AI sector, potentially driving innovation and reducing costs for enterprises that rely on AI solutions. The open-source nature of GLM 5.2 allows for broader collaboration and faster iteration, which could accelerate the pace of AI development globally. For U.S. companies, this may necessitate a reevaluation of their strategies to maintain competitiveness, particularly in light of the cost implications highlighted by the introduction of OpenAI's Jalapeño chip.
What's Next?
As Zhipu's GLM 5.2 continues to gain traction, U.S. companies may need to adapt by embracing open-source models or enhancing their proprietary technologies to maintain a competitive edge. The ongoing inference-cost race, influenced by new technologies like the Jalapeño chip, will likely intensify, prompting companies like Nvidia and Broadcom to innovate further. Additionally, the broader adoption of open-source AI models could lead to increased collaboration across industries, potentially reshaping the landscape of AI development and application.
