Redesign: The Art of Iteration

The Daily Interesting — Art & Design Edition

February 1, 2026

The idea

“Good design is redesign. It’s rare to get things right the first time. Experts expect to throw away some early work. They plan for plans to change.” — Paul Graham, “Taste for Makers.” (via BarnsworthBurning) (barnsworthburning.net)

Why it’s interesting

Iteration isn’t a lack of vision; it’s the vision maturing in public. Treating drafts as sacrificial layers lets you approach clarity without clinging to first impulses. In art and design, this turns perfectionism into practice: you ship, learn, and revise, knowing the real piece emerges through cycles. (barnsworthburning.net)


Who said it (and the context)

Paul Graham is a programmer, essayist, and cofounder of Y Combinator. In “Taste for Makers,” he distills patterns of good design across art, products, and architecture—arguing for timelessness, simplicity, and the courage to keep refining. The line above sits alongside his observations that “good design looks easy” and that the best work often comes after multiple passes. (barnsworthburning.net)


A small expansion

Iteration benefits from a target. “Redesign” isn’t endless churn; it’s a sequence of informed bets that narrow toward a clearer whole. Two pragmatic cues help:

  • Keep “throwaway” cheap and visible: rough studies, cardboard models, low‑fidelity screens.
  • Define what you’re optimizing toward: legibility, affordance, atmosphere—so each pass has a job.

Think of the work as a conversation: mark, step back, respond. Over time, the piece teaches you how to finish it. (barnsworthburning.net)


Another lens on the idea

Industrial designer Raymond Loewy’s MAYA principle—Most Advanced Yet Acceptable—offers a compass for iterative decisions: push toward what’s new, but land where audiences can still recognize the thing. Research on typicality vs. novelty backs this balance. Used with Graham’s “redesign,” MAYA helps you decide when a revision is better versus merely stranger. (oreilly.com)


Why this matters across art and design

  • Visual identity

    • Build systems that invite revision: alternate logo locks, optical sizes, and motion rules. Publish versioned specs so changes read as care, not drift. Aim revisions at legibility and memorability, not novelty alone. (barnsworthburning.net)
  • Product & UX

    • Prototype the “wrong” thing quickly to learn the right thing. Use staged fidelity: sketches → clickable wireframes → instrumented betas. Tune toward MAYA: introduce one advanced move per release, keep the rest comfortably typical. (oreilly.com)
  • Architecture

    • Mock life with full‑scale, temporary elements—tape walls, foam blocks, light tests—then redesign details where circulation or comfort falter. Alternate close‑up fixes with far‑view judgments so the whole stays coherent. (barnsworthburning.net)
  • Studio art

    • Treat series as laboratories. Iterate formats (scale, substrate, sequence) while holding a constant (palette, motif). Scrap, sand, repaint—let traces of earlier states inform the final surface. (barnsworthburning.net)

Two quick drills

  1. The One‑Variable Sprint (30 minutes)
  • Take a current piece and generate five variations changing only one variable: weight, spacing, stroke order, tempo, or light direction.
  • Pick the strongest change and fold it back into the base. Repeat once. You’ve “advanced,” yet stayed “acceptable.” (oreilly.com)
  1. Kill‑and‑Keep Crit (25 minutes)
  • Print or pin up three iterations. For each, circle one element to kill and one to keep.
  • The next pass must remove all “kills” and preserve all “keeps,” forcing decisive redesign instead of vague tweaks. (barnsworthburning.net)

Thread to remember

We’ve studied failure as teacher and walking as a thinking tool. Today’s lesson ties them: progress comes from paced oscillation—advance, assess, advance. Design proposes, you revise; the audience recognizes, you push a little further. Redesign isn’t backtracking. It’s how the work finds its best form. (barnsworthburning.net)


Sources

  • Central idea and excerpts: Paul Graham, “Taste for Makers.” (via BarnsworthBurning). (barnsworthburning.net)
  • Creative process vantage shifts: Frank Chimero, The Shape of Design. (barnsworthburning.net)
  • Complementary perspective: MAYA (Most Advanced Yet Acceptable), Universal Principles of Design; overview of Loewy’s concept. (oreilly.com)
  • Background on MAYA attribution: German Wikipedia entry. (de.wikipedia.org)

—
Tip: Name your next file “v0.3” on purpose. It reminds you the real work is still ahead—and that’s the point.