AI Design Tools Research
A six-month research initiative mapping how designers actually use AI tools — what they adopt, what they quietly abandon, and why the gap exists.
The Question
AI tools for designers are proliferating faster than research can keep pace. Every week brings a new claim: AI will replace designers, AI will empower designers, AI changes nothing. I wanted to find out what was actually happening in studios.
Over six months I conducted 34 interviews with designers across disciplines — brand, product, motion, type, spatial — and observed their workflows directly. The research was self-funded and deliberately independent of any tool vendor.
What I Found
The headline finding is unsexy but important: adoption is high, continued use is low. Most designers try AI tools quickly and enthusiastically. Most stop using them within two months. Not because the tools are bad — because they solve problems designers don't have.
The specific friction points, in order of frequency:
-
Output fidelity to intent. AI tools are excellent at generating options but poor at generating the right options. Designers found that curating AI output took longer than making the work themselves for anything requiring genuine taste.
-
Context loss. Current tools have no memory of a project's history. Every prompt starts from zero. Senior designers in particular found this intolerable — their process is deeply contextual.
-
Credit and ownership. When a designer cannot explain how an artifact was made, they feel uneasy showing it. This is not a legal concern (mostly). It is an identity concern.
Implications
The tools that will persist in design workflows are not the generative ones — they are the reasoning ones. Tools that help designers think through a problem, not tools that produce an artifact at the end of the problem.
The report was shared with several tool-builders and has informed the direction of two product roadmaps (under NDA). A condensed version was published as a long-form essay.