AI 时代的设计：品味、工艺与观点

Design in the Age of AI: Taste, Craft, and Point of View

迪伦·菲尔德 Dylan Field · 杨彼得 · 2026-04-12 · 约 41 分钟 · 原视频 ↗

本期速览 · Overview

Figma CEO Dylan Field 探讨设计在 AI 时代不断演变的角色，强调品味、工艺和观点的重要性。

Figma CEO Dylan Field discusses the evolving role of design, emphasizing the importance of taste, craft, and point of view in the age of AI.

品味、手艺和观点是 AI 增强设计中不可替代的人类技能。
Taste, craft, and point of view remain irreplaceable human skills in AI-augmented design.
直接画布操作比提示更精准地调整设计。
Direct canvas manipulation beats prompting for precise design adjustments.
设计正成为新的代码，实现视觉优先的拉取请求到生产。
Design is becoming the new code, enabling visual-first pull requests to production.

品味是驾驭可能性并有清晰偏好；工艺是在所有层面打磨细节；观点是表达独特洞察。
Taste is navigating possibilities with clear preferences; craft is pushing details at all levels; point of view expresses a unique insight.
AI 输出是起点而非终点；人类的迭代和判断至关重要。
AI outputs are starting points, not final results; human iteration and judgment are essential.
在画布上直接操作优于通过提示词调整设计。
Direct manipulation on a canvas is superior to prompting for design adjustments.
领导者应亲自构建产品以激励团队并创造转折点。
Leaders should build things themselves to inspire teams and create inflection points.
设计正成为新的代码；以视觉为先的设计将能直接提交生产级拉取请求。
Design is becoming the new code; visual-first design will enable pull requests to production.
行业仍处于 AI 整合的早期；发散-收敛循环是创造力的关键。
The industry is at the very start of AI integration; divergence-convergence loops are key for creativity.

AI 不会取代工作；人类的品味和工艺对最终输出仍然至关重要。
AI won't replace jobs; human taste and craft remain critical for final output.
设计系统和自动布局虽有好处，但可能限制初学者的创造力。
Design systems and auto layout can constrain beginners' creativity despite their benefits.
尽管谈论角色融合，我们看到更多的是职责融合而非角色融合。
Despite talk of role merging, we see more merging of responsibilities, not roles.
顶尖工程师现在才开始看到 AI 在编码中的效用；它尚未全面超越人类。
Very good engineers are only now starting to see AI's utility in coding; it's not universally superior yet.
最好的产品拥有并非所有人都认同的观点。
The best products have a point of view that not everyone agrees with.
AI 能帮助探索发散的可能性，但人类仍是判断何为优秀的裁判。
AI can help explore divergent possibilities, but the human is still the judge of what's good.