Design by Prompt: The Monolithic Dilution of Creativity

Bish, bash, bosh, slop. I've been wondering recently where the design industry is going, where it is at the moment and, perhaps more importantly, where I hope it will eventually settle. With the prevalence of AI-generated work, it's easy to feel slightly disillusioned by it all, and to wonder whether we're in the process of losing something far larger than we realise.

Most designers, whether they work in graphics, interiors, fashion, photography, architecture, products or landscapes, didn't enter the industry because they enjoyed operating software. They did it because they enjoyed creating things, and I was no different. At 11 years old I'd designed every feature and undulation of an 18-hole golf course that I fully intended to build when I was older (currently on hold). I designed cars that could fly and buildings with increasingly elaborate features that Bruce Wayne would be proud to own. Most of it was completely ridiculous, obviously, but I loved designing things, and so it was fairly inevitable that I ended up pursuing a career in design.

That's probably why what's happening in the industry now makes me feel a little uneasy. We've adopted AI into some areas of our own workflow at BakerWilcox, so this isn't an argument for pretending it doesn't exist, refusing to use it or bashing it on the head. I can see first-hand how effective it is at creating multiple ideas in seconds, how quickly those ideas can be developed and how polished, realistic and viable the results can be. For many clients, that's going to be enormously appealing.

But design has always been about the chase, the dead ends, the false starts and that occasional moment when an idea arrives from nowhere and somehow solves a problem you've been wrestling with for days. That's the creative process. The assumption with AI seems to be that creativity is the outcome: the logo, the advert, the campaign or the visual. I've never really seen it that way. Creativity is the journey that gets you there, and for most of my career the best ideas haven't arrived fully formed. They've emerged somewhere between the first terrible thought and the twentieth, occasionally during a conversation that had absolutely nothing to do with the brief, sometimes after a walk and sometimes after a pint, but rarely on command.

What feels different now is that the industry increasingly celebrates the answer whilst quietly removing the process that produced it. Type a prompt, generate fifty concepts, refine the best three and present the winner. It's undeniably efficient, but creativity was never really about producing options, and designers have always been able to do that. What clients are paying for, whether they realise it or not, is judgement and the ability to recognise the difference between something that is merely attractive and something that actually means something, answers the brief and gets us to where we want to be positioned.

The trouble is that AI-generated design often looks remarkably good. The lighting is perfect, the composition is perfect and increasingly the typography is strong too. Yet after looking at enough of it, you begin to notice a strange sameness. The outputs aren't identical, of course, but there are recognisable markers that begin to tie them together, and this is where I think it starts getting interesting for all of us.

The promise of design has always been differentiation. Different designers arrive at different answers, and that's one of the wonderful things about creativity. To an untrained eye you might reasonably say, "Well, AI has just given me fifty different ideas to choose from", and on the face of it, it has. But look a little harder and you begin to wonder just how different and, more importantly, appropriate those fifty ideas really are.

The problem is that the untrained eye is often attached to the person commissioning the work in the first place. Why call in the designers when we can do this ourselves? It's easy to understand the appeal, because cost and speed are two fairly substantial juggernauts when somebody is making a commercial decision. If something looks good, arrives in seconds and costs next to nothing, you'd need a fairly convincing reason not to use it.

Extend that thought across everything in our lives touched by design and it becomes much more interesting. Frank Lloyd Wright, Dieter Rams, Charles and Ray Eames, Coco Chanel, Jony Ive, Le Corbusier, Paul Rand and Norman Foster all created work that influenced far more than the object, building or graphic sitting in front of them. Their ideas travelled, with other designers studying them, challenging them, borrowing from them and taking some of that thinking into completely different projects of their own. That's how creative culture develops. One idea doesn't simply get repeated; somebody else gets hold of it and pushes it somewhere new.

If all of that can eventually be diluted into a prompt asking AI to "create something like this for me", where are we headed? My fear is that we end up in some kind of Noddy land where every created touchpoint has been designed, and probably made, by a machine that thinks it knows what we want. The really worrying bit is that we'll probably think we want it too, because if nobody is pushing the boundaries anymore, eventually we forget where the boundaries were.

Businesses come to agencies because they want to stand apart from their competitors, to be noticed, remembered and occupy a space nobody else occupies. Yet many of those same businesses are now turning towards systems trained on enormous libraries of existing work, which is a slightly curious strategy. If your competitive advantage comes from being different, relying entirely on a machine that has learnt from everything that already exists isn't the most obvious place to go looking for originality.

The uncomfortable truth is that a lot of design jobs will disappear, not necessarily because AI is producing better work, but because it is producing work that is good enough. For many businesses, good enough at a fraction of the cost is a perfectly rational decision, and in truth some projects probably never needed a designer in the first place. The danger comes when businesses start confusing efficiency with originality, particularly when the work they're saving money on is the very thing that's supposed to give them an edge over everyone else.

At BakerWilcox we've always believed that the real value of design sits in the idea, in the thinking, the unexpected connection, the insight nobody saw coming and the smoking gun concept that suddenly makes everything else feel obvious. That's the part I worry about losing, because an industry built on creativity appears increasingly comfortable removing the creative process from creativity itself. Once you've done that, you're left with an awful lot of output, some of it good, some of it bad, most of it perfectly acceptable and very little of it unforgettable.

There is, however, a more positive way this could play out, and I suspect we're already beginning to see it. AI will settle down and find its place. At the moment AI-generated content is everywhere, to the extent that "AI slop" has already entered everyday language, but the sheer ease with which it can be produced may ultimately be the thing that reduces its value. When a global brand and someone's grandma can both generate the same result in seconds, polish alone stops being particularly interesting.

Apple's recent identity work for Apple TV offers an interesting counterpoint. Rather than simulating the effect digitally, its creative team made physical versions of the logo from glass and experimented with real light, movement and cameras to create the finished ident. It would undoubtedly have been possible to produce something convincing on a computer, and probably considerably easier, but the process itself mattered. There was craft involved, experimentation and people making decisions about how something should behave in the real world. That's where the value lies, because somebody has thought about how the brand needs to behave and, just as importantly, how it needs to be seen to behave.

That's where I hope we eventually land with AI. Brands will need to find a place for it, and designers will too, but hopefully not as a replacement for creative people. Used within a creative process and directed by people with ideas, judgement and experience, it could become an extraordinary tool. Used as a replacement for the process itself, I worry that we'll simply produce more and more things that look perfectly good and say absolutely nothing. Bish, bash, bosh, slop.

The world still needs outstanding design. It needs remarkable thinkers and creatives producing buildings, photography, fashion, furniture, products, identities and ideas we haven't seen before. What it doesn't need is Jason from accounts deciding that "we can just ask ChatGPT".