设计未来:Brian Lovin 谈 AI、Notion 与角色融合
Designing the Future: Brian Lovin on AI, Notion, and Blurring Roles
布莱恩·洛文 Brian Lovin · Dive Club · 2026-04-14 · 约 55 分钟 · 原视频 ↗
打开互动全文版(中英对照 + 朗读 + 问答)→
本期速览 · Overview
Brian Lovin 分享了设计师如何拥抱适应性与 AI,结合他在 Notion 的紧张入职经历和 AI 代理工作。
Brian Lovin shares how designers must embrace adaptability and AI, drawing from his intense onboarding at Notion and work on AI agents.
要点 · TL;DR
- AI 将设计师的角色从视觉设计转变为构建基于代码的框架和原型。
AI changes the designer's role from visual design to building code-based harnesses and prototypes. - 理解技术系统和精确的提示词使 AI 成为强大工具,而非魔法。
Understanding technical systems and precise prompts makes AI a powerful tool, not magic. - 在 AI 炒作和可能的倦怠中,专注力和适应力是关键的竞争优势。
Focus and adaptability are key competitive advantages amid AI hype and potential burnout.
核心观点 · Key points
- 设计师应该在代码中做原型并设计智能体的“缰绳”,而不只是用 Figma。
Designers should prototype in code and design the harness for agents, not just Figma. - 设计、产品管理和工程的边界正在模糊;专注于交付优质软件,而不是头衔。
Design, PM, and engineering boundaries are blurring; focus on shipping great software, not titles. - AI 的“缰绳”每六个月就会变一次;保持适应并理解背后的原因。
AI harnesses change every six months; stay adaptable and understand why. - 高质量的提示词和技术词汇能显著提升 AI 的输出效果。
High-quality prompts and technical vocabulary significantly improve AI output. - AI 不是魔法;理解底层计算机系统能让你更高效。
AI is not magic; understanding underlying computer systems makes you more effective. - 在混乱的 AI 时代,不被干扰的专注能力是一种竞争优势。
The ability to focus without distraction is a competitive advantage in the chaotic AI era.
反共识 · Contrarian takes
- 尽管 AI 被大肆炒作,模型在最终视觉打磨上仍然糟糕;手工编辑代码是必要的。
Despite AI hype, models are still bad at final visual polish; hand-editing code is necessary. - 「没人写代码」的说法是假的;阅读和编写代码仍是质量的保证。
The “nobody writes code” meta is false; reading and writing code still ensures quality. - 自我报告的 AI 生产力提升可能是虚幻的;产出更多但质量更低、修复更多。
Self-reported AI productivity gains may be illusory; more output but lower quality and more fixes. - 预计 AI 倦怠和回落即将到来;不可持续的高强度会导致撞墙。
Expect an AI burnout and come-down soon; unsustainable intensity leads to hitting a wall. - 会编程能让你在用 AI 时获得巨大优势,因为你知道正确的词汇。
Knowing how to code gives a major advantage with AI because you know the right vocabulary. - 不针对 AI 可读生态系统进行重新定位的公司,有被淘汰的风险。
Companies that don't reposition for an AI-legible ecosystem risk becoming obsolete.
本期章节 · Chapters(共 15)
- 拥抱持续变化 Embracing constant change
- 原型游乐场保真度 Prototype playground fidelity
- 生产中的设计师 Designers in production
- 协作转变 Collaboration shifts
- 工作流假设反转 Workflow assumptions reversed
- 处理边缘情况 Handling edge cases
- 快速报告和修复漏洞 Fast bug reporting and fixing
- 岩石抛光机类比 The rock tumbler analogy
- AI提升上限 AI raises the ceiling
- 苏睿作为解药 Shuri as an antidote
- 周计划周末执行 Weekly planning, weekend execution
- 东京设计大会FOMO Tokyo Design Conference FOMO
- AI不是魔法 AI Is Not Magic
- Notion文化与增长 Notion Culture and Growth
- 最后的问题 Final Question
阅读全文双语转录 →