DeepSeek's User Drop: What's Fueling the AI Battle Outside the Box?

Exploring the Shifting Sands of AI Competition and the Rise of Open-Source Solutions

Welcome to this edition of our newsletter, where we delve into the dynamic shifts within the AI landscape. As competition heats up and new innovations emerge, we invite you to ponder this: How will the decline in DeepSeek's user retention reshape the future of AI, and what opportunities may arise from these challenges in the realm of open-source technologies?

🚀 AI Shake-Up Alert

• DeepSeek's user retention has faced a significant decline, with a 150-day post-launch report indicating a sharp drop, raising concerns about its position in the AI landscape.
• A notable shift towards open-source models is evident, with a 20-fold increase in their usage, reshaping competition in the AI service space globally.
• At the Fifth China Integrated Circuit Design Innovation Conference, Shenzhen Zhongxing Microelectronics highlighted the potential of RISC-V architecture to enhance AI computational efficiency, addressing the demands of large language models and further supporting advancements in drug discovery and healthcare AI applications.
• Read more on: AI领域新动向：具身智能融资激增与DeepSeek流量下滑的双重影响
• Explore the advancements in RISC-V for AI technology here: 中兴微电子RISC-V架构推动AI计算效率提升的最新进展

🧠 Insights Unpacked

• Why RISC-V matters in AI: The RISC-V architecture is emerging as a pivotal player in enhancing AI computational efficiency, particularly vital for large language models (LLMs). As highlighted during the recent Fifth China Integrated Circuit Design Innovation Conference, RISC-V offers an open, flexible, and scalable framework that could address the growing demand for computational power in AI applications (source: 中兴微电子RISC-V架构推动AI计算效率提升的最新进展). This is especially crucial given the limitations of traditional GPU solutions, which often fall short due to high costs and specialized applications.

• Collaboration key for chip designers: The effective evolution of RISC-V relies heavily on collaboration among chip manufacturers, developers, industry users, and research institutions. This collective effort aims to standardize instruction set extensions and enhance toolchain maturity, positioning RISC-V as a critical engine for AI technological advances. Companies like Shenzhen Zhongxing Microelectronics are spearheading these efforts, calling for an industry-wide cooperation to define technical standards, which is essential for accelerating the deployment of AI technologies.

• Discover the future: For deeper insights into how RISC-V architecture can shape the landscape of AI and support advancements in drug discovery and healthcare AI applications, explore the full article here: AI领域新动向：具身智能融资激增与DeepSeek流量下滑的双重影响.

🌟 Intelligence Revolution

• Embodied intelligence gains traction: In 2025, the embodied intelligence sector has experienced a remarkable funding boom, as illustrated by Xinghai Tu successfully raising over $100 million during its A4/A5 funding rounds. This investment surge has more than tripled its valuation since the beginning of the year, demonstrating a significant milestone in the commercialization of embodied intelligence technologies.

• Pharma breakthroughs with AI: Google DeepMind's Isomorphic Labs is on the verge of achieving critical milestones in drug discovery, particularly with their AI-assisted cancer drug development project. This initiative showcases the transformative potential of AI in the pharmaceutical industry, aiming to pave the way for innovative treatments and medical advancements.

• What this means for healthcare: The lowering of barriers in AI application development, propelled by advancements in open-source models and contextual engineering, is facilitating greater innovation within the pharmaceutical and healthcare sectors. As noted in the recent newsletter, such trends are allowing more developers and companies to engage with AI technology, ultimately enhancing problem-solving capabilities in high-value applications.

• Question for creatives: 'How will the surge in embodied intelligence funding transform your next move?'

For more insights on the dynamics of funding in embodied intelligence and the advancements in AI for drug discovery, refer to the articles AI领域新动向：具身智能融资激增与DeepSeek流量下滑的双重影响 and 中兴微电子RISC-V架构推动AI计算效率提升的最新进展.