How AI Changed My Design Workflow, Not My Standards

A Tool, Not a Shortcut

When AI tools started showing up in real creative workflows, I had the same skepticism a lot of designers did. Would this replace thinking with prompting? Would output start looking generic, fast, and a little soulless?

A year of actually using it in client work later, my answer is: only if you let it.

What AI actually replaced for me

Not the ideas. Not the direction. Not the judgment calls about what a piece needs to say or how it should feel.

What it replaced was the slow, mechanical parts: generating variations quickly to test a direction, producing base imagery I'd then refine and composite by hand, exploring options I wouldn't have had time to explore manually under a tight deadline.

Where the human part still matters, maybe more than before

Here's what I noticed almost immediately: AI-assisted output that skips human judgment is easy to spot. It has a certain sameness to it, ironically, the exact problem I wrote about a few weeks ago.

The designer's job didn't shrink. It shifted. Less time spent on manual production of every single option, more time spent curating, refining, and making sure what ships still feels intentional, not generated.

My actual rule for using it

If a client can tell an asset was AI-assisted, and not in a good way, I didn't do my job. The tool should disappear into the output. What should remain visible is judgment: knowing what to keep, what to change, and what needs a human hand to actually land.

AI gave me more room to think. It didn't do the thinking for me. That distinction is the whole difference between AI as a shortcut and AI as a genuine part of the craft.