Ask HN: 为什么营销软件还没有迎来它的“光标时刻”？
1 分•作者: richbelt•5 个月前
在构建了两年的人工智能营销工具之后，我深信，大多数人工智能应用，用 Pete Koomen 的话说，都是“无马的马车”——它们将人工智能硬生生地嫁接到现有的界面上，而不是构建原生的人工智能体验。
营销问题在于：营销人员为了一个活动，需要在 5-8 个工具之间来回切换。比如，用 Hootsuite 进行日程安排，用 Canva 进行设计，用 ChatGPT 生成文案，用 Analytics 获取洞察。这种切换带来的认知负担会扼杀生产力和创造力。
传统的营销软件迫使你学习多个界面。我们正在尝试一种不同的方法：通过对话动态生成界面。你不用再去“社交媒体日程安排”界面，而是说“给我创建一个关于最新人工智能趋势的 X 话题”，然后相应的用户界面就会出现，显示帖子的预览和发布功能。类似的理念也适用于其他营销用例。
技术方法：1/ 意图识别，用于解析对话请求；2/ 基于任务需求的动态组件组装；3/ 跨对话线程的持久状态管理；4/ 用于多平台操作的集成编排。
假设：在人工智能时代，界面应该适应用户的意图，而不是迫使用户去适应静态界面。
像 Cursor 和 Lovable 这样的应用程序之所以如此出色，是因为它们从一开始就以人工智能为核心构建，并充分利用了它们的对话和生成能力。我们设想为营销也采用类似的方法。
如果你经常在 Twitter、Reddit 等社交媒体上发帖，并且有兴趣测试一种新颖的营销软件方法来帮助我们验证这个概念，我很乐意与你交流。
查看原文
After building AI marketing tools for 2 years, I've become convinced most AI applications are, in the words of Pete Koomen, "horseless carriages" - they bolt AI onto existing interfaces instead of building AI-native experiences.<p>The marketing problem: Marketers context-switch between 5-8 tools for a single campaign. Hootsuite for scheduling, Canva for design, ChatGPT for copy, Analytics for insights. The cognitive overhead of switching kills productivity and creative flow.<p>Traditional marketing software forces you to learn multiple interfaces. We're experimenting with a different approach: conversation that dynamically generates interfaces. Instead of navigating to the "social media scheduler," you say "build me a X thread on the latest AI trends" and the appropriate UI appears displaying a preview of the post and publishing capabilities. Similar idea across other marketing use cases.<p>Technical approach: 1/ Intent recognition to parse conversational requests, 2/ Dynamic component assembly based on task requirements. 3/ Persistent state management across conversation threads. 4/ Integration orchestration for multi-platform actions<p>The hypothesis: In the AI era, the interface should adapt to the user's intent, not force users to adapt to static interfaces.<p>Apps like Cursor and Lovable work so well because they were built with AI at the heart of it and leaned into their conversational and generative capabilities. We’re envisioning a similar approach for marketing.<p>If you're active online posting across social media sites like Twitter, Reddit, etc. and are interested in testing a novel approach to marketing software to help us validate this concept, I'd love to talk to you.