Topics
September 24, 2026
September 24, 2026
As more people can make what once required a specialist, it’s tempting to ask what design still owns. Taste, judgement, curiosity and craft are often offered as answers. They matter, but none belongs to designers alone.
Design has never been static. Throughout my career, disciplines have emerged, separated, recombined and been renamed. Teams have centralised and decentralised. We have moved from documents to shared files, from pages to systems, from static artefacts to products that continuously evolve. Change itself is familiar. What feels different now is the compression. Ideas that once took weeks to render can increasingly become tangible in hours, often through the efforts of people who would previously have needed considerable specialist support.
For many designers, that’s exhilarating. It’s also understandably unsettling, particularly if we have defined our value through the work only we could produce. But claiming exclusive ownership of a set of qualities seems unlikely to help us understand what is changing. The more useful question may be what those qualities allow us to do, and what responsibilities come with them when more people can shape an experience.
I find myself returning to the work of interpreting a situation, recognising what matters, and helping an idea survive the competing demands of making it real. That work has always involved other people’s expertise. As the distance between imagining something and producing it shrinks, our ability to listen, challenge and make considered decisions together may matter more than any boundary we try to preserve around the discipline.
Over time, every experienced practitioner develops something resembling a personal operating system. It’s difficult to capture faithfully in a playbook: an accumulation of instincts, principles, experiences, scars, preferences and questions that shape how we interpret what is in front of us.
Mine has changed considerably. Earlier in my career, a problem would often trigger an almost immediate visual response.
That instinct has never disappeared, and I hope it never does. I remain fascinated by the ability of experience and visual communication to turn an otherwise ordinary interaction into something people feel connected to.
There is something remarkable about taking an activity as unremarkable as sending someone money, ordering a car or buying clothes and making it feel coherent, distinctive, or even enjoyable. We often reduce this to usability or aesthetics, but exceptional design frequently does something less easily quantified. It makes an ordinary commercial exchange feel considered. Perhaps, in rather unsophisticated language, it makes something cool: the kind of cool that comes from somebody caring enough to see an overlooked experience differently.
As my career has progressed, however, the initial creative question has accumulated company.
Moving from individual creator to team member, team leader and eventually organisational leader has meant becoming responsible for progressively more perspectives without wanting to lose the creative sensibility that made the work interesting in the first place. I suspect that tension is familiar to many design leaders. We spend less time making the work ourselves and more time creating the conditions in which good work can happen. Remaining close enough to recognise whether those conditions are working becomes part of the responsibility.
This is partly why I’ve become increasingly interested in systems. At their best, design systems:
A strong brand system does something similar, providing enough context that hundreds or thousands of people can make decisions that remain recognisably connected to the same organisation.
There is an important tension within any good system, though: we create rules because we want people to follow them, and we need talented people who know when to question them.
A mature designer interprets a principle against the situation in front of them. They understand its intention, recognise when the context has changed and, occasionally, create something better than the system anticipated. Treating every departure as drift would prevent the system from learning. Accepting every departure without scrutiny would gradually make it meaningless.
Design leadership has to govern that relationship between consistency and change. Some decisions need to remain dependable so that teams can build on them. Others need room for curiosity, interrogation and play. The difficulty lies in knowing which is which, particularly when a proposed improvement benefits one product but creates complexity for everyone else.
That requires judgement. No sufficiently interesting organisation can encode every possible future decision in advance, and a system that removes the need to think is unlikely to remain useful for very long.
Inside commercial organisations, there are very few absolutes available to a design leader. Holding uncompromisingly to every design position might sound principled, but it is unlikely to make you particularly effective. Organisations are built through relationships, negotiation and competing constraints. Continually pushing back without understanding the context around you eventually erodes the trust required to influence anything at all.
The job involves fairly representing the expertise of the discipline while remaining capable of hearing the expertise surrounding it. That includes knowing when something important is being compromised and when the idea is developing through contact with perspectives it needs.
One experience leading a large design-system programme reinforced this for me. The original ambition was relatively straightforward: consolidate fragmented systems and substantially raise the quality of the experience. The difficult part was creating the conditions required to get there.
The proposition needed to make sense commercially and for the brand. Designers needed to believe in ambition. Engineers needed to see themselves as participants rather than recipients. An undertaking that initially carried the urgency of a design project had to acquire the patience required of an organisational programme.
One conviction remained particularly important: improving quality could not be achieved solely by producing better components in a design file. The distance between design and engineering itself had to be reduced. That made the work broader than the original consolidation task. It also meant accepting that the route towards better quality would have to be shaped with the people responsible for delivering it.
The distinction mattered. The quality ambition remained, while the scope and ownership of the work evolved. What initially required significant alignment eventually became a broader shared mission because the conversation moved beyond what the system looked like towards what it could enable.
I see stewardship in that willingness to let the means evolve while remaining responsible for the ambition. A design system can be beautifully resolved and still leave the organisation struggling to deliver a better experience. The leadership task extends into that gap.
The prospect of agents participating in this work makes these questions more consequential. There is a temptation to imagine an agentic future as a more perfect version of automation: define the system correctly, document the rules comprehensively, and allow increasingly capable machines to execute faithfully against them.
I suspect the reality will be more nuanced. Agents will encounter situations in which instructions are incomplete, principles pull in different directions or a familiar rule produces an unsuitable result. We will need to distinguish what is authoritative from what permits interpretation, and decide when variation is an error or a reason to reconsider the system itself.
We already navigate versions of these questions with people. We expect experienced practitioners to understand a system and contribute to its evolution. That comparison has limits. Human and machine judgement are not equivalent, and a plausible explanation from an agent is not sufficient evidence that its decision is sound.
For me, the organisational implication is that permission to act needs to come with clarity about who reviews the consequences. One way to make that discussion concrete is to take a workflow in which an agent will participate and identify its decisions:
There is a practical tension here. Requiring a person to approve every variation could consume much of the time the technology saves. Allowing variation without meaningful review could multiply inconsistency and rework. The appropriate balance depends on the consequences of being wrong, and somebody has to remain accountable for choosing it.
The question therefore extends beyond whether an agent can follow our rules. It includes whether we have created an environment that can respond sensibly when those rules are insufficient. Increased autonomy makes the quality of that environment—and the judgement behind it—more important.
Perhaps that is where I currently land. I’m wary of making confident claims about some permanent boundary between human and machine when that boundary is moving so quickly. What I do think remains important is the responsibility to interpret.
That means:
Design is relational. Products of any meaningful complexity require people to negotiate decisions under pressure, with incomplete information and consequences they cannot fully anticipate. Expertise matters within that negotiation, but so does the willingness to let somebody else’s expertise alter your own understanding.
Perhaps that is why thoughtful curiosity still matters so much to me. It is the willingness to look again, especially when every measurable requirement appears to have been satisfied and something still feels wrong. It helps us recognise when a constraint deserves respect and when it deserves a question.
The tools will keep changing, and the organisational structures probably will too. More people will be able to participate in shaping an experience. Our systems will increasingly include agents, and the distance between imagining something and making it real will continue to collapse.
I don’t think the answer is to protect design from any of this. It’s to become clearer about what we were trying to protect in the first place: the responsibility to interpret, challenge and steward an idea through complexity, while preserving enough of what made it worth pursuing.
For now, at least, that is what I find myself holding onto.
.jpg)
