Get less generic design out of AI

Written by AI

TLDR: Anshu Chimala, ex-Apple AI design lead, outlines a Discover-Define-Deliver process pushing AI past generic outputs using seed strings, ambitious prompts, and subagent design critics.
Best for: Owners building landing pages with AI who keep getting the same template.
The tip: Have the AI generate a random string and base its colors, layout and type on it. Then have a second agent critique screenshots until it scores the page 9/10 or higher.
What you need: Claude Code
an OpenAI or Gemini API key, for images
The prompt: Build me a landing page for [business]. First generate a long random string of letters and numbers with a script, then choose the colors, layout and typography from patterns you see in it. Keep the string out of the design.
our prompt
Start here: Have the AI seed each design with a random string.
Give it an ambitious, specific creative brief.
Have a second agent critique screenshots, not code, and cut anything without a job.
Filed note: Anshu Chimala spent twelve years prototyping AI products at Apple; this Lenny's guest post lays out his end-to-end process for getting real design taste out of AI. The full walkthrough sits behind the paywall.

Highlights: Large language models are next-token predictors: at each step, they look at a sequence of text and predict what comes next based on millions of examples.
Great design, on the other hand, starts with feeling and aims to create an emotional response. It bends the rules and delights users with memorable, unexpected choices.
One technique for this is String Seed of Thought, published by Sakana AI. We make the AI generate a random string and use it as design inspiration.
AI loves to add more, but it rarely takes away. One of the biggest signs that a design is AI-generated is that it overexplains everything or contains elements that don’t serve any practical purpose.
Excerpt: I’d always thought AI was bad at design. But after reading this mind-blowing post by Anshu Chimala, I realize I was just doing it wrong. Anshu led software engineering and design teams at Apple for 12 years, focusing on research and prototyping for future AI products.
From lennysnewsletter.com