What if the most valuable design skill of 2026 isn't mastering new AI tools — it's being willing to make something ugly?
The great flattening
The conversation about design homogenisation isn't new. We've been watching brands shed their distinctiveness for years — swapping characterful logos for clean sans-serifs, trading bold colour palettes for muted pastels, smoothing every rough edge until everything looks like it rolled off the same Figma template.
But AI has put its foot on the accelerator.
The problem is structural. Generative AI tools are trained on existing visual culture — which is already converging. So when you ask an AI to generate a brand identity, a hero image, or a layout, it pulls from a pool of aesthetics that were already starting to look alike. The output is competent. It's polished. And it's indistinguishable from ten thousand other outputs generated by ten thousand other designers asking similar prompts.
This isn't subtle any more. AI doesn't just reflect the "blanding" trend. It amplifies it. Every prompt that defaults to "modern," "clean," and "minimal" pushes the visual centre of gravity further toward sameness. The industry is finally naming the problem clearly — and it's about time.
Why sameness is your opportunity
Here's where I'd normally be expected to wring my hands about the death of creativity. I'm not going to do that.
Because when everything looks the same, distinctiveness becomes the scarcest — and most valuable — thing a designer can offer.
Think about it from a client's perspective. If AI can produce a perfectly adequate, perfectly generic brand identity in an afternoon, then "perfectly adequate" is now the baseline. It's table stakes. The thing clients will actually pay for — the thing that moves the needle for their business — is work that doesn't look like everything else.
That's not an AI skill. That's a human one.
The case for grit
There's a telling example in game design right now. The team behind Silent Planet is deliberately drawing on the raw, textured aesthetic of 1990s visual culture — not as shallow nostalgia, but as a conscious rejection of hyper-refined contemporary polish. They're embracing imperfection, tactile roughness, and visual noise in an era that's drowning in smoothness.
What's interesting isn't the retro angle. It's the underlying impulse: to choose something that feels made by a human hand.
This isn't just a game design trend. It's a signal. Audiences — and clients — are starting to feel the sameness. They may not be able to articulate it yet, but they respond when something breaks the pattern. They stop scrolling. They remember the brand. That's worth more than any amount of algorithmic polish.
Your anti-reference library
Here's a practical move you can make this week: start building an anti-reference library.
Most of us collect references — mood boards full of work we admire, styles we want to emulate. An anti-reference library is the opposite. It's a collection of the visual defaults you want to actively avoid. The same gradient blobs. The same geometric sans-serif. The same pastel-on-white landing page. The same AI-generated hero illustration with soft lighting and no discernible style.
Pin them up. Study them. Understand why they're everywhere — because they're safe, they're fast, they're what the algorithm suggests. And then make deliberate choices to go somewhere else.
This doesn't mean being weird for weirdness's sake. It means developing a point of view. Choosing a typeface because it has character, not because it's the first suggestion. Adding texture because the brief calls for warmth, not because grain is trendy. Making a colour choice that feels specific to this brand, not to "brands in general."
The real skill gap
There's a lot of anxiety right now about whether AI will replace designers. I think that's the wrong question.
The better question: will AI replace designers who only produce work that looks like AI output?
Probably, yes.
But designers who bring genuine taste, cultural awareness, and the confidence to push a brief somewhere unexpected? They're about to become more valuable, not less. AI has commoditised competence. It hasn't commoditised point of view.
So basically…
AI is accelerating design homogenisation — not because the tools are bad, but because they're trained on a visual culture that was already converging. The output is safe, polished, and forgettable.
Your advantage isn't in prompting better. It's in seeing differently. Build an anti-reference library. Embrace texture, imperfection, and specificity. Develop a point of view that no model can replicate.
The bland middle is getting more crowded by the day. The edges are wide open.