my notes from "AI X Design: (how) will AI change how we design?" event by DeVc
1/ With all the advancements in AI, the role of a designer remains the same = Biz Problems + User problems + User Behaviour.
2/ There were four generations of design tools:
- Gen 1: Photoshop, CorelDRAW, Illustrator. Focus was on static images, vector tools, bitmap/vector manipulation.
- Gen 2: Sketch, InVision. Focus was on multi-screen interfaces, UI kits, component libraries.
- Gen 3: Figma, Adobe XD. Focus was on multiplayer design, plugins, collaboration, and developer handoff.
- Gen 4: Niti AI, Lovable, Claude, ChatGPT. Focus is on AI-assisted design, multi-agent systems, and enhanced creativity tools.
3/ With previous tools, a lot of time used to go into learning the tool β now itβs going to critical thinking.
4/ LLMs thrive on patterns, but great design (or art) happens outside the patterns. This makes the role of a designer even more valuable.
5/ Design systems exist to enable visual consistency AND reusable code.
6/ Best way to use GPT for graphic work:
Create a few graphics β Reverse engineer prompt for those images β Generalize the prompt β Create new graphics.
7/ I was very impressed by Harish (Head of Design and Marketing at CRED) speaking about Figma MCP. I always believed, similar to vibe coding, vibe designing or vibe environment design or vibe game designing will emerge, and MCPs will play a critical role in doing this.
8/ A fascinating AI use case I have heard of is synthetic user testing. First, you train your model on the immediate challenges, motivations, ICP data, and more. Then you give designs and ask GPT to assess the output.