Key takeaways
- Design is shifting from a visual layer to a business function measured by outcomes.
- Design system debt is invisible until it stalls velocity—fight it with ownership, audits, and deliberate deprecation.
- AI agents are now users too; build interfaces that are legible to machines, not just human eyes.
- Perceived 'aliveness' comes from designed behaviour and trust signals, not decoration.
- Open, psychologically safe culture is the operating system your design system runs on.
The most expensive word in modern product teams is design—not because good design costs money, but because bad systems quietly bleed it. A design system built without governance becomes debt. An interface built for humans alone becomes obsolete the moment an AI agent tries to use it. And a design team treated as a decoration service never gets a seat at the table where the real decisions are made. Scaling design well means confronting all three problems at once.
Design as a function, not a finish
The shift we keep seeing across serious product organisations is philosophical before it's practical: design is being reframed as a function of the business rather than a visual layer applied at the end. Pixels are the output; the value is in the decisions—what to build, for whom, and why. When leadership treats design as a function, designers stop asking "how should this screen look?" and start asking "what outcome does this need to produce, and how do we measure it?"
This matters commercially. Teams that anchor design to function can defend roadmap choices with logic instead of taste, and they win the arguments that used to be lost to "just make it pop." It also changes hiring, tooling, and how you justify investment to a CFO who cares about conversion, retention, and cost-to-serve—not gradients.
The quiet cost of design system debt
Every design system starts as a promise: consistency, speed, less rework. Most accumulate design system debt instead—orphaned components, three buttons that should be one, tokens nobody trusts, documentation that lies. The debt is invisible on any dashboard until velocity stalls and every new feature triggers a debate about which pattern is canonical.
Fighting it is unglamorous but decidable:
- Assign ownership. A system with no maintainer is a graveyard with good intentions. Someone must be accountable for deprecation, not just creation.
- Audit on a cadence. Quarterly component audits catch drift before it calcifies into a maintenance tax.
- Deprecate loudly. Removing patterns is harder than adding them—and more valuable. A smaller, trusted system beats a sprawling, ignored one.
- Measure adoption, not existence. A component nobody uses is debt, not an asset.
The tooling landscape is loosening the old constraints too. Builders like Kirki, bringing an infinite freeform canvas to WordPress, hint at a future where design isn't trapped inside rigid boxes and plugin sprawl. Freedom is welcome—but freedom without a system is just faster debt. The discipline has to scale alongside the canvas.
Designing for AI, not just for eyes
The next constraint on design is that humans are no longer the only users. AI agents now read, parse, and act on interfaces. Emerging principles for AI interaction push designers to think about legibility for machines: clear structure, semantic meaning, predictable states, and explicit affordances. An interface that only communicates through visual cues—colour, position, animation—is illegible to an agent that can't "see" the way we assume.
There's a subtler lesson here about perception. The reason an animated character or a chatty assistant feels alive isn't the face—it's the behaviour we design around it. Animacy is triggered, not drawn. As AI moves deeper into products, the design job becomes orchestrating behaviour and trust signals, not decorating a mascot. Design for what the system does under uncertainty, how it recovers from error, how it explains itself. That's where confidence is won or lost.
Culture is the system behind the system
None of this scales without an open design culture, and you don't need a leadership title to build one. Psychological safety is what lets a junior designer flag that a component is broken, or challenge a pattern before it ships to a million users. The teams that catch debt early and adapt fastest to AI-native interfaces are the ones where people feel safe to say "this is wrong" out loud.
Practically, that means sharing work early and unfinished, framing critique around outcomes rather than ego, and rewarding the person who prevents a mistake as much as the one who ships a feature. Culture is the operating system your design system runs on—neglect it and every process fix eventually corrodes.
Where this leaves your team
The organisations that win the next cycle won't be the ones with the prettiest interfaces. They'll be the ones treating design as a function tied to business outcomes, paying down system debt deliberately, building interfaces legible to both humans and agents, and protecting the culture that makes all of it possible. That's the shape of design that actually scales.
Sources
- The face was never the point — UX Collective
- Design system debt, game UX, 39 principles for AI interaction — UX Collective
- Design as a function — UX Collective
- Meet Kirki: WordPress’s First Visual Builder With An Infinite Canvas — Smashing Magazine
- How any designer can build a more open design culture at work — UX Collective
