阶跃星辰AI终端 · 智谱摸高战略
阶跃星辰AI终端 · 智谱摸高战略
一、 权威必看
EN: The emergence of AI phones marks a critical shift in the hardware industry, moving from simple smart devices to “super carriers” of intelligence. Jieyue Xingchen is set to release its first AI agent phone, manufactured by Huaqin Technology, which goes beyond simple OEM to include deep binding cooperation. This partnership leverages the Step-GUI open protocol, allowing developers to deploy Agent capabilities on phones in as little as 10 minutes. The collaboration between ByteDance and ZTE Nubia for the 2026 World Artificial Intelligence Conference (WAIC) further validates this trend.
中： 手机正成为大模型竞争的新赛场，从智能硬件向“超级载体”转型。阶跃星辰即将发布首款AI智能体手机，由华勤技术代工并深度绑定合作，而非简单贴牌。基于Step-GUI开放协议，开发者可在10分钟内部署具备Agent能力的手机。此前中兴努比亚Z80 Ultra已搭载该模型，实现跨应用调用工具。今年6月豆包手机已验证AI与系统深度绑定的可行性，通过系统将App串联。在即将举行的2026世界人工智能大会上，字节跳动联合中兴努比亚打造的首款AI智能体手机将亮相，展示AI终端的最新进展。
二、 深度与多元
EN: The core concern for the second half of the year is not index fluctuations but global liquidity redistribution. AI companies like OpenAI, Anthropic, Google, and Microsoft are increasing capital expenditures at unprecedented speeds, acting as the new “water pumps” that draw wealth from traditional sectors like real estate and internet platforms. This shift means that while investment in AI surges, liquidity in other areas may dry up, creating a stark contrast between asset prices and actual cash flow availability.
中： 下半年真正需要担心的不是指数涨跌，而是全球流动性的重新分配。过去二十年，房地产和互联网平台曾是财富的“抽水机”，如今这一角色正被AI巨头接替。OpenAI、Anthropic、Google和微软等公司以前所未有的速度增加资本开支，导致资金从传统领域向AI基础设施集中。这种趋势下，虽然AI投资规模巨大，但其他领域的流动性可能枯竭。投资者需警惕这种结构性变化，理解资本流向背后的逻辑，而非仅仅关注表面新闻。
三、 科技与财经
EN: Zhipu AI, listed on the Hong Kong Stock Exchange, has announced a two-year strategic investment plan named “TouchHigh.” Founder Tang Jie emphasized that the company will not pursue short-term monetization but will focus entirely on achieving AGI. The plan targets four key areas: long-horizon task capabilities, next-generation memory architecture, autonomous reasoning, and multimodal perception. This decision reflects a commitment to foundational research despite market pressures for quick returns.
中： 港股大模型巨头智谱（02513.HK）创始人唐杰发布内部信，启动名为“TouchHigh”的两年战略投入计划。公司明确表示不追求短期应用变现，全力冲击下一代通用人工智能（AGI）。投入将集中在四大方向：攻坚长程任务能力、研发新一代记忆架构、自主拆解跨年级任务、以及多模态感知。唐杰回顾智谱从2006年学术搜索系统到2021年GLM-130B的研发历程，强调“别人敲钟，我们归零”的信念。这一战略表明智谱在港股上市后选择全面回归基础模型研究，以长期主义对抗行业浮躁。
四、 国际视野
EN: On the sports front, Erling Haaland’s performance has drawn widespread discussion on social media, with netizens criticizing his “dreamwalking” in a recent match. While this is a sports topic, it reflects the global attention to high-profile athletes and the intense scrutiny they face. The viral nature of such discussions highlights how digital platforms amplify individual performances into global narratives.
中： 哈兰德在近期比赛中的表现引发微博热搜热议，网民批评其“全场梦游”。虽然这是体育话题，但反映了全球对顶级运动员的关注及舆论压力。数字平台将个人表现放大为全球叙事，体现了社交媒体时代的传播特征。此类讨论虽非时政核心，但展示了公众情绪的快速发酵机制，可作为观察社会心理的微观窗口。
【传闻隔离 / 网络讨论】 关于哈兰德表现的批评属于网民主观评价，并非官方定论或技术事实，请理性看待体育竞技中的正常波动。
五、 青年与生活
EN: Developers face technical challenges in building AI agents, particularly with OpenAI’s tool_calls.arguments being strings rather than objects due to streaming architecture compromises. A developer on Juejin shared a four-step parsing pipeline that resolved three crashes over two months, offering a practical solution for handling JSON parsing errors in Agent development.
中： 开发者在构建AI智能体时面临技术挑战，OpenAI的tool_calls.arguments为字符串而非对象，导致JSON.parse频繁出错。掘金博主拆解SDK源码后提出四步解析管道，两个月零崩溃，解决了这一痛点。该分享针对流式架构下的数据格式问题，提供了可复用的代码思路。这反映了青年开发者在AI落地过程中的务实探索，通过技术细节优化提升系统稳定性。
【21ZHAO 综合判断】
EN: The convergence of AI hardware, liquidity shifts, and foundational research strategies reveals a clear macro trend: capital and talent are concentrating on AGI infrastructure. For developers, this means focusing on robust data handling and long-term model capabilities rather than short-term hacks.
- Prioritize building resilient parsing pipelines for Agent interactions to handle streaming data formats effectively.
- Monitor liquidity trends in AI investments to anticipate market shifts and adjust technical roadmaps accordingly.
中： AI硬件、流动性转移与基础研究的交汇揭示了宏观趋势：资本与人才正集中投向AGI基础设施。对开发者而言，这意味着应关注数据处理的鲁棒性与模型的长期能力，而非短期技巧。
- 构建健壮的Agent交互解析管道，以有效处理流式数据格式，避免JSON解析崩溃。
- 监控AI投资领域的流动性变化，预判市场趋势并调整技术路线图，确保项目与资本流向同步。