Our AI Elephant Needs Attention

How AI is troubling for design -- but not for the reasons you might think

Andrew Boardman, in a warm room in Montreal, writing a newsletter on a laptop, surrounded by three black cats. There is an elephant in the room. (I appreciate the AI making me ten pounds lighter and 10 years younger.) Credit: DALL-E.

Dear Designer:

At some point, when the dust settles some, I do want to talk about the election results in the U.S. last week. We should focus on what may happen politically in the coming years to designers, especially independent ones and those focused on mission, art and social impact.

But first, I wanted to dig into something a tiny bit deeper, because, after all, isn’t this like my first real issue of Dear Designer? And why would I start with something small like autocracy, censorship and reformation?

Today’s topic is AI, every designer’s favorite elephant in the room.

I watched an intriguing presentation last week that was part of RGD’s Design Thinker’s conference in Toronto. Robert Wong, VP of Google Creative Lab, gave a buoyant talk about their take on where design is heading in the midst of the AI “revolution”.

He demonstrated a number of fantastical use cases for AI, including one in which Google asked four illustrator-artists (Eric Hu, Shawna X, Erik Carter and Haruko Hayakawa) to offer their ideas and various stylistic ideations to illustrate the original Alice in Wonderland text.

The resulting art project, called Infinite Wonderland, also “employed” John Tenniel, the English illustrator of the 1865 version of Alice’s Adventures in Wonderland, who drew nearly 100 individual drawings. I grew up staring at those wild illustrations. They had — and still have — such a strange grip on me. Do you remember them? The Caterpillar-Man smoking a hookah, that Cheshire Cat’s creepy wide grin, Alice always looking angry and agitated, the Mad Hatter who had that oversized price tag (‘In this style 10/6’) stuck to the side of his lid.

All of them infinitely wonderful.

In interacting with Google’s Infinite Wonderland site, you can watch images being generated one by one in real time by the processing algorithm, each styled according to the likes of Tenniel or any of the four contemporary illustrators. They are mesmerizing and oddly engaging.

Cascading Style Shots

But they are not infinitely wonderful.

We need to remember that these are also only styled images. The drawings themselves are simply outputs from a machine being told by humans how to manneristically present a given subject. They are surface expressions by a visual language learning machine that is built to entertain and decorate the page.

What these images are not doing, what they cannot yet do, is say something truly meaningful about the original content. Yes, some of the images are “originals” from the artists, who had to provide working comps in order for the Google AI to learn. Mostly, these images are mere interpolations, indirect assignations of visual ideas that are styled to look like they came out of the wet brains of human artists. These generated images are simply simulations — or simulations of simulations.

Take a look.

Prompt: Alice falling head first down a rabbit-hole, she is looking up at a signpost pointing to the left.

Same prompt as above.

The images are kind of interesting but they can’t hold a candle to those original drawings from 1865 by Tenniel. Compare.

The Caterpillar, from John Tenniel's illustrations from the first edition of Alice's Adventures in Wonderland (1865). Credit: Wikipedia

What is my point? I agree with Wong who makes the oft too heard claim that AI is just a tool. Yes, it’s a pencil, it’s a camera, it’s a canvas. It’s a plaything with which humans can tinker and trial. The AI itself is a more sophisticated way for us to bring human ideas into an existing form in a more expedient and inexpensive way.

Visually generative artificial intelligence — whether via DALL-E, Google Labs, Midjourney, Firefly, etc — is the medium. It is not the message. That’s the pitch of Big Tech.

But that’s not true. For our technology overlords, AI is also the message. AI is in service to all of humanity. AI will do good things for all of us eventually. AI will love us if we love it back. AI will make our work better, faster, and stronger. Corporations are expanding human creativity and offering it back to us for free or for a low, low cost. AI is just the latest milestone in the history of human imagination.

But there are two big problems with this message. First, AI offers a false narrative that these corporations care about creativity and those who create. They don’t — they mostly care about market share. That’s why so much money is swashing around right now. Second, AI is delivering work built on the backs of millions of artists, designers, writers and coders that have voluntarily and freely added to the open source treasury of knowledge we call the internet.

As VR pioneer Jaron Lanier said over ten years ago: “Digital information is really just people in disguise.”

At the end of his talk, Wong was asked whether the four artists were paid for their work and whether Google acknowledges and values the vast treasury of visual images to which it is indebted. He proudly noted that all four of the artists were paid for this project, going on to say that Google pays for its right to use the internet as a vacuum, something I do not see in its online statement about AI ethics.

Was Tenniel or his heirs paid? When DALL-E made the image above of me writing in my lair with an elephant and three cats in the room, who exactly was compensated? Yes, it’s a terribly terrific picture and I don’t mind it.

But who gains here? Is it a zero sum game? Will all of us benefit from AI? Or will a handful Big Tech companies gain from AI, while the rest of scavenge and surrender?

I’m not against AI. I do see it as a potentially incredibly useful design tool. It also might one day cure cancer.

What if AI companies generated, along with each image, code, or copy, a micro transaction to compensate creators upon every prompt? Could scholarships, grants or foundations that support basic income emerge that supports designers and artists? How can we hold these Big Tech accountable to what they are taking and to what they are offering in return?

The problem is the message.

Last January, musician Nick Cave was asked about ChatGPT and its ability to write a song “in the style of Nick Cave”. His response: “What makes a great song great is not its close resemblance to a recognizable work. Writing a good song is not mimicry, or replication, or pastiche, it is the opposite. It is an act of self-murder that destroys all one has strived to produce in the past.”

See you next week, dear designer.

Yours,
Andrew

---

Image of the Week

Design by Glenn Goluska (1947–2011) for the book, El Lissitzky, The Topology of Typography – imprimerie dromadaire, 1983. Credit: Canadian Typographic Archives

Russian artist and designer El Lissitzky (1890–1941) left a profound mark on markmaking. For Lissitzky, letters are not letters and forms are not forms — they are elements of “pure feeling”, expressions of what we cannot see and cannot know. Above is a page designed by Canadian designer Glenn Goluska (in the style of Lissitzky) from his 1983 book about El Lissitsky, one of the founders of Suprematism and abstract art.

---

Quote of the Week

AI will probably most likely lead to the end of the world, but in the meantime, there'll be great companies.

~ Sam Altman, CEO of Open AI (makers of ChatGPT and DALL-E)

---

If you’ve been forwarded this email, you can get your very own sub right here. See you soon.