Designers Don’t Need AI to Design
We want tools that deepen research, improve flow, and respect and advance our craft
First, a big old welcome to the dozens of new subscribers to Dear Designer. I’m tickled and honoured that you’re here.
I hope you like what you read and see.
If you don’t, let me know. If you do, let me know. I’m really open to any feedback. Here goes.
Dear Designer,
I’m reading both (and multiple) sides of the AI conversation and the complexity is maddening and deep. The clouds move in and then they move out, the light in hot pursuit of putting heat on the subject. The air is heavy.
One global analyst shows that we are heading toward economic devastation and a time of untold social depravity. Another presents a future as one that, while imperfect, presents us lifted from the rigours and ridiculousness of a life of modern task management. And yet another shows that we will enter a period of high dalliance, happiness and pleasure.
For us designers, I think we should start the question with a therapeutic one.
What is it that we want?
What do designers want from our chosen path and our imagined future? If we start from here — from desire and hope — we have the opportunity to interrogate what generative AI and LLMs offer and discover what we might want to learn from them. In contrast, starting from need is what the AI corporate overlords are asking of us. They, for the most part, are addressing corporate, shareholder, and market needs, framed by and connected to revenue, profit, and audience share, respectively.
Two caveats before we jump in. First, despite having a darker outlook as a younger man, I now jump into a conversation about the present moment with less focus on human precarity, which is indeed real and getting more frightening, and more focus on human potentiality. I don’t see the world through rose coloured glasses, but what is shocking to me, right now, is that we could be on the precipice of creating a more humane, sustainable, peaceful, stable and unhurried world of work — if, the right mix of state, organizational and bureaucratic actors were in place.
(Rutger Bregman, who in 2019 berated the billionaires in Davos to “stop talking about philanthropy and start talking about taxes” apparently aligns here. I’m interested in reading his new book called Moral Ambition, which makes a case for people to create legacy careers around societal benefit instead of professional ones that build personal wealth.)
Second, while I’m only one designer with a particularly digital and social democrat bent, I am taking a stab at answering with the hope that it will open up a conversation around what it is that you, dear designer, need from generative AI, LLMs and other non-human intelligence models.
Finally, there are those of you will say, and rightfully so, that you don’t need AI and that you want nothing to do with it. That AI is wasteful, resource intensive, grotesquely non-private and job killing.
I sympathize 100%. The problem with a fully anti-AI stance is that the horse left the barn 10 years ago and the buggy is still in there. AI may not be as revolutionary as some indicate, it may take more time for it to deeply affect knowledge work, and it may be more nefarious than we can imagine. But it will change us. And grappling with our needs now is one pragmatic way to address what it is we theoretically want.
Preliminary thoughts on what designers need from the AI
Accurate research. To a large extent, what I want from AI is what it already delivers — the capacity to help me read, research, and review ideas related to a given topic. I am building a small website of quotes from famous, infamous and uncelebrated people and I am using a bit of AI to help me not only find relevant quotes but ensure that my sources are accurate. It works well enough but the AI requires constant fact checking and rechecking on my part. I don’t trust half the links it provides — and the other half — well, they are about as accurate as general search results. The AI offers a good textual foundation for the limited investigation and exploration I’m doing. But for richer and more sophisticated visual research, AI is a long way from being realizable, relevant or reliable. Google Image search has become contaminated with poor facsimiles of imagery from non-cited sources and the large AI models don’t yet feature the capacity to dig into the image goldmine that is the web. What do I need here? Accuracy and depth.
Better visitorship. For digital design projects — I’m mostly speaking about the web — the one goal for all of our clients is to ensure that the relevant folks visit the relevant parts of a given website. Designers currently use a set of heuristics, user journeys, best practices, testing, heat maps and other models to help match what we want and expect audiences to see first, second and third on a given site or app. Plus guesswork. And finger in the wind. And unlimited doses of hope. What we need is to better be able to assure a client that a visitor is going to get the right details right the first or second time to a given project. If AI can help make educated predictions around user journey strategy, I believe that clients and their audiences succeed.
Greater collaboration. I don’t know what this means, actually. Saving…
Supporting surprise. The most memorable design offers up a challenge, a way of surprising the mind, creating a memory or cognitive conflict that propels the recipient to take a new breath or take action. This may be reducing 10,000 words into 10. Or using a single and outrageous colour. Or a terrible font. Or something more grandiose or humble or grotesque or contrived. Even for traditional of media, audiences seek to be entertained, engrossed, energized. A big headline, a weird photo, a strange caption, a goofy title, an upended quote, an awful animation. If there is one thing that we will increasingly appreciate, it is humour. Hilarity is what drives drivel on most social channels these days. In the AI’s hands, I can see it going all kinds of wrong. But as the models learn irony and those inscrutable human qualities called mordancy, sarcasm and pique, perhaps the AI will offer up suggestions to surprise us, making our work more human, motivating and meaningful.
More time. This is what all of us really want. You can have all of the money in the world but if your life is short (and it is), money is not supremely meaningful. Time to play. Time with our friends. Time to sit. Time to eat, walk, run, frolic and f*ck up. Generative AI’s capacity to offer increased productivity is what is being touted. But for designers, what does that mean? For me, I want the AI to help me like a work pal might. Prioritize my tasks for me and remind me when I forgot something. Tell me when I am missing something or misremembering. Check my work. Offer me guidance on how I can make something better. Give me a hint or two without doing it for me.
Essentially, I want the AI to be my virtual assistant, art director, and energetic visual researcher to be there sitting beside me, readied, and reliable.
What do designers not want? We don’t want AI to freaking design for us. We don’t want it to “make design” as Figma excitedly presents now. We don’t want it to find crappy fonts from shitty foundries. We don’t want its fake imagery to show up in our designs.
We also don’t want the AI to narrate our presentations. And we definitely don’t want the robots to fake-tell us that we’re doing great, thank you very much.
If we want anything from AI, we want it to support our work like a good work pal might — and then go away, so that we, too, can go home and make our art.
Yours,
Image of the week
The great California-based foundry Emigre has released a 1,264-page chronicle of all of its type specimen books between 1986 and 2024. It’s a massive accomplishment and the book — Emigre Fonts: Type Specimens, 1986–2024 — represents a (heavy) cornerstone in the history of design. I have a number of Emigre Magazines and type specimens but having them all bound together is pretty cool. Produced through Letterform Archive, I think this might be added to my collection soon.
Quote of the week
The precise role of the artist, then, is to illuminate that darkness, blaze roads through that vast forest, so that we will not, in all our doing, lose sight of its purpose, which is, after all, to make the world a more human dwelling place.
~ James Baldwin
Thank you for reading, dear designer. Would you do me a favour and share this with a designer/artist friend? My goal is to get to 1,000 subscribers, after which I will consider another goal.
Also, if you want to subscribe or recommend this thing, you can do that here.