The design community spent this week piling onto a restaurant owner for using an AI-generated logo. And sure, there are real conversations to be had about AI and creative work. But here's what's been nagging me: most of us still aren't testing whether our own designs work when someone increases their font size. Which one actually hurts users more?
What happened
Two stories landed on my radar this week that, put side by side, say something uncomfortable about where we focus our energy.
First: a restaurant owner named Rachel Smith faced such intense backlash over her AI-generated logo that she was forced to replace it. Designers and the wider creative community called it out as devaluing professional design work. The response was fierce enough that Smith spoke publicly about how overwhelming the whole experience was.
Second — and much more quietly — a detailed walkthrough emerged showing how Figma variables can be used to test font size scaling as part of your normal design workflow. It's a technique that makes one of the most fundamental accessibility requirements almost automatic to check. No separate audit. No afterthought. Just… baked into how you already work.
Guess which story got more attention.
The uncomfortable contrast
I'm not here to defend AI-generated logos. If you're building a brand, investing in a designer who understands your story, your audience, and your market is almost always worth it. The AI output might look passable at a glance, but it won't hold up under the kind of scrutiny a real identity needs to survive.
But let's be honest about what happened here. A small business owner — not a multinational, not a tech giant — made a budget decision, and the design community responded with enough collective fury to force a change. We can spot an AI-generated logo from across the room. We've trained our eyes for it.
Meanwhile, how many of us can say with confidence that our last project handled font scaling properly? That someone who needs 200% text zoom can actually use the thing we shipped?
I'd wager the number is embarrassingly low.
Accessibility failures affect real people every single day. They're not theoretical. They're not about professional pride or industry economics. They're about whether someone with low vision can read a menu, complete a form, or navigate a page. And yet accessibility rarely generates the same passion — the same outrage — that AI does in design circles.
Why accessibility keeps losing the attention war
It's not hard to understand why. AI feels like an existential threat to how we make a living. It's emotional. It's personal. Accessibility, by contrast, feels like homework. It's important and we all know it, but it doesn't trigger the same fight-or-flight response.
The problem is that this mismatch in energy has consequences. We're brilliant at policing other people's design decisions — especially when those decisions feel like they threaten our profession. We're much less brilliant at holding ourselves to the standards that actually matter to users.
And that's where the font scaling thing comes in.
Tool spotlight: font scaling with Figma variables
Testing font scaling has traditionally been a pain. You either do it in the browser after handoff (too late to catch layout issues cheaply) or you manually resize everything in your design file (tedious enough that nobody does it consistently).
The approach using Figma variables changes that. By setting up your type sizes as variables with multiple modes — one for default, one for scaled — you can toggle between them instantly across your entire file. Want to see what your layout looks like at 150% or 200% font size? One click.
It's not a new Figma feature, exactly. It's a clever use of the variables system that's been there for a while. But the insight is in making the accessibility check feel like a natural part of the design process rather than a separate task. And that distinction — built-in versus bolted-on — is everything when it comes to whether a team actually does it.
If you work in Figma and you haven't set this up yet, it's worth the thirty minutes it'll take. You'll catch layout-breaking issues earlier, and you'll ship designs that work for more people. That's not a nice-to-have. That's the job.
So basically…
We're very good at spotting when someone else cuts a corner that threatens our industry. We're much less good at addressing the corners we cut ourselves — the ones that affect the people using what we build.
I'm not saying the AI logo conversation doesn't matter. It does. But if we're going to position ourselves as the people who care about quality, craft, and doing things properly, then that standard has to apply to our own work first.
Before you join the next pile-on, open your last shipped project and bump the font size to 200%. See what breaks. That's where the real craft lives — and it's where users actually need us to show up.