AI is making design easier.

Good.

“We do hard things” (thanks JD) is a mantra I’ve been repeating to the Maven team ad nauseam for a few years now. Partly in response to the endless ads promising that the latest bit of tech will make everything effortless. Partly because that's what interests us and keeps us sharp.

We’re early adopters of technology. We understand where it can remove repetitive work, speed up our workflow and give us more time to think.

But we also understand the danger.

The danger comes when easy starts shaping the answer.

When the thing that’s easiest to generate becomes the thing that gets made. When the first plausible answer starts looking like the right answer. When everything slowly gravitates towards the large language model median.

Polished. Plausible. Perfectly fine.

Design has always had this problem. Committees have it. Focus groups have it. Competitor moodboards have it. “Can we make it a bit more like them?” has it. We’ve spent our careers battling the gravitational pull of the middle and protecting good ideas while everyone, usually with the best intentions, tries to knock their difficult edges off.

There’s a quote by Brian Wilson, co-founder of the Beach Boys, that I love:

“Beware the lollipop of mediocrity;
lick it once and you’ll suck forever.”

Crude, but useful.

We do hard things

The interesting thing is that the hardest parts of design were never really the making.

Sometimes the hard thing is zigging while an entire industry is zagging. Sometimes it’s standing in a confusing building and working out how a first-time visitor can find reception without thinking about it.

It’s getting architects, contractors, AV teams, developers, marketers, suppliers and clients pulling in the same direction when everyone has different priorities.

It’s taking something complicated and making it feel simple. It’s knowing when not to design something. Knowing when the rough edge is the interesting bit and should be left alone. Knowing when there are too many ideas, too many messages, too many signs, too many pages or too many features – and having the confidence to take things away.

That reductive phase of design is often the hardest bit. Adding is easy. Removing requires judgement.

And increasingly, our work is about relationships.

The relationship between a customer and a business. Between an employee and the company they work for. Between a visitor and a space. Between what a brand promises and what actually happens when someone encounters it.

Sometimes fixing that means designing a new identity, website, campaign, space or wayfinding system. Sometimes it means finding the awkward problem underneath and fixing that instead.

That is usually harder than papering over the cracks with something shiny.

Design isn't art

Design is not self-expression. It is the practical act of helping someone else solve a problem through clear thinking, good judgement and skilled execution. The visual result matters, of course. Craft matters. Taste matters. But those things are part of the answer, not the job in its entirety.

That is why the claim that AI is “killing design” feels too shallow. AI is becoming very good at producing things that look designed, and that will change the industry. Some production work will shrink. Some will disappear. But producing the visible artefact was never the whole job.

A large part of design happens before anything is made at all: understanding the real problem, spotting the gap between what a business says and what people actually experience, deciding what needs changing and, just as importantly, what does not.

We’ve been rethinking Maven’s own language for this reason. The word design has become increasingly associated with aesthetic-first, surface-level output. That has never really described what we do. It is simply the most visible part of the work.

The harder part is making sense of complexity, reducing it to something clear, getting people aligned around the right idea and then making that idea real without sanding all the character out of it. AI can help us do plenty of that faster. We intend to use it wherever it makes the work better.

But easy shouldn’t decide the answer. So if AI can make the easy things easier, brilliant. That gives us more time to do the hard things.