HBO Max embraces vertical video with a new 'Shorts' feed HBO Max拥抱竖屏视频，推出全新‘短片’频道

TL;DR

• HBO Max introduces a TikTok-like vertical video feed to enhance content discovery, leveraging AI-powered scene analysis and user-specific preferences.
• The platform launches an experimental conversational search feature using natural language understanding to recommend movies and TV shows based on open-ended queries.
• This move aligns with broader industry trends where streaming services adopt short-form video feeds and AI-driven search to improve user engagement and simplify content navigation.

Why It Matters

This development highlights the growing importance of AI in enhancing user experience within streaming platforms, particularly in addressing content overload and improving discovery mechanisms. For AI practitioners and researchers, it underscores the practical applications of natural language processing and recommendation systems in real-world consumer products. Additionally, it reflects a competitive landscape where streaming services are increasingly adopting similar features to retain and engage users.

Technical Details

• Vertical Video Feed: The new "Shorts" icon provides a customizable feed of trailers, clips, and bonus content, tailored to individual watch history and preferences. AI tools analyze thousands of hours of film and TV show content using scene-level metadata to surface compelling clips, which are then curated by HBO Max editors.
• Conversational Search: This feature employs natural language understanding to interpret user queries such as "in the mood for a comedy" or "best movie for a girls night in," recommending relevant titles. It is currently available to select Android users in the U.S., with plans for expansion.
• Comparative Analysis: Similar features have been introduced by other major streaming platforms like Netflix (AI-powered search), Disney+, and Peacock (vertical video feeds), indicating a trend toward integrating advanced AI technologies to streamline content discovery.

Industry Insight

Streaming services are increasingly relying on AI to differentiate themselves in a crowded market, focusing on personalized experiences and intuitive search functionalities. As audiences become more accustomed to short-form content on platforms like TikTok, incorporating vertical video feeds can help maintain user engagement and reduce churn. For AI professionals, this presents opportunities to explore advancements in natural language processing, recommendation algorithms, and user behavior modeling to further optimize these features.

TL;DR

• HBO Max 推出类似 TikTok 的短视频流（Shorts），基于用户观看历史和偏好进行个性化推荐，帮助用户发现内容。
• 引入 AI 驱动的自然语言对话式搜索功能，支持如“想看点喜剧”或“适合女生夜聊的电影”等模糊查询。
• 短视频内容通过 AI 分析电影/剧集的场景级元数据生成精选片段，由编辑团队最终筛选呈现。
• 该功能已在部分美国 iOS 和 Android 用户中测试，计划逐步扩展至更多设备和市场。
• 此举标志着流媒体平台在内容发现机制上全面转向短视频化与 AI 语义理解，以应对观众注意力碎片化趋势。

为什么值得看

本文揭示了主流流媒体平台如何通过融合短视频形态与AI语义搜索重构内容发现路径，对娱乐行业数字化转型具有标杆意义。对于AI从业者而言，其场景级元数据分析与自然语言推荐系统提供了可复用的工程范式，尤其在非结构化影视内容处理方面具备参考价值。

技术解析

• 短视频推荐引擎：利用用户历史行为数据训练个性化模型，结合场景级元数据（如情绪、角色关系、镜头类型）从数千小时素材中自动提取高吸引力片段，经人工审核后投放至垂直视频流。
• 对话式搜索模块：基于NLU（自然语言理解）技术解析口语化查询意图，映射到内容标签库实现跨维度匹配（如“ dysfunctional family drama ”→家庭冲突+剧情类+成人向），并支持上下文连贯交互。
• 多模态内容索引体系：构建包含视觉特征、音频节奏、剧本结构等多维度的影视内容图谱，为AI剪辑与语义检索提供底层支撑。
• 灰度发布策略：采用分阶段 rollout 模式，先在小范围特定设备（iOS/Android）验证稳定性与用户体验，再根据反馈迭代优化算法参数与界面逻辑。
• 竞品对标机制：明确将Netflix的AI搜索与Amazon Fire TV语音助手作为参照系，强调差异化在于更细粒度的场景切片能力与自然语言自由度。

行业启示

• 流媒体竞争焦点正从“内容储备量”转向“内容触达效率”，短视频+AI组合成为降低用户决策成本的标准配置，传统导航菜单将被动态信息流取代。
• 影视制作端需提前布局“可拆解性”创作——即在设计阶段就考虑哪些场景更适合被AI识别为独立高光时刻，以适应未来自动化分发需求。
• AI应用落地应注重“人机协同”而非完全替代，当前案例中编辑团队仍保留最终裁量权，这既保障了品牌调性控制，也为算法偏见修正留出缓冲空间。