近日，好未来发布了截至2025年5月31日的2026财年Q1财报，其Q1净收入从上年同期的4.14亿美元上升 到本季度的5.75亿美元，亮眼成绩的取得在于学习服务和AI驱动的智能硬件稳步增长。事实上，作为国 内教育大模型的"头号玩家"，AI已经融入好未来的各大业务中，各项产品已经被AI深度重构。 T4学习机的AI助手正引导一名孩子逐步拆解复杂的数学难题；而几步之遥，另一名小观众则沉浸在素 养机器人的拼搭挑战中——在上海世博展览馆内学而思展台上围满了前来体验的小观众们。随着以大模 型技术为代表的人工智能加速发展，教育正成为AI垂类应用落地最快的领域之一。 AI战略开花结果 2026财年Q1，好未来净收入从上年同期的4.14亿美元上升到本季度的5.75亿美元，归属于好未来的净利 润为3128.2万美元，同比增长174.36%。好未来的利润率也持续上涨，而这背后是好未来规模效应释 放、运营流程优化以及AI技术提效。 所谓AI技术提效，是好未来在AI大模型上的重点布局的三条路径。即AI学习教练，利用AI Agent来协助 老师完成诸如跟踪学生出勤情况、制定学习计划、回答问题以及管理日常事务等任务，提高了工作效 率；A ...

Group 1 - The 2025 World Artificial Intelligence Conference highlighted a shift from theoretical discussions about large models to practical applications in various industries [3] - DeepSeek is recognized as a pivotal product that enhances reasoning capabilities in education, particularly in programming and solving mathematical problems [3] - The education sector is increasingly focused on the problem-solving abilities of large models, with significant advancements noted in AI's reasoning capabilities [3] Group 2 - Kingsoft Office launched the WPS Lingxi intelligent product, allowing users to create documents and presentations through natural language and multi-turn dialogue [4] - The industry remains focused on foundational models rather than vertical industry models, which are more about applications and solutions than the models themselves [4] - Kingsoft Office's assistant president expressed caution regarding the maturity of Agent technology, noting that it has not yet fully addressed core user pain points [4] Group 3 - Guangyi Mingdao, a digital technology company under Geely Holding Group, reported significant challenges in deploying large models in industrial AI due to the proprietary nature of factory data and the disconnect between process knowledge and AI technology [4][5] - Many manufacturing companies are eager to adopt AI but struggle to integrate it into their business processes effectively, leading to a disconnect between AI capabilities and operational execution [5] - The lack of deep understanding of industrial processes limits the potential impact of AI on manufacturing enterprises, despite the perceived intelligence of large models [5]