022: AI-generative technologies
[During - Session 7] Speed. The Bell Curve. What words mean. Emotions. Also, can you draw a fish riding a bicycle? And the Great Wall of Art has cracks in it, finally.
Would you like an Executive Summary?
I’m trying something new here at the top. If you like it, you can vote at the end of today’s message and motivate me to continue. (TL;DR = Internet for “Too Long; Didn’t Read”—also Internet for “Executive Summary.”)
TL;DR
AI-generative’s rapid pace is definitive. As Benedict Evans wrote, “One of the most striking things about the current wave of generative machine learning is that the breakthrough technology has been followed by a wave of actual companies and products much, much more quickly than in the past.”
You want to get on board sooner this time. People could ignore social for years, and the Internet for longer. AI-generative has real utility you can leverage, right now. Those with fluency will lead.
This isn’t a movie, or comic book. We’ve been picturing the idea of AI longer than we’ve actually engaged with AI. That’s where most of the cultural tension exists—misrepresenting sentience, “intelligence,” and intent.
The prompt could be mightier than the pen, or the paint brush. We’re about to unlock all kinds of pragmatic creativity. Old tropes around who gets to make art will disappear.
Tonight was the seventh session of The Future of Advertising. We’re half way through the curriculum. We introduced the fourth major theme: Artificial Intelligence. Also, it snowed about four inches today in Minneapolis, all of it wet and heavy. Big ups to the E-Go 24in electric snow thrower.
If this is your first time here, welcome! This is a “During” post, recapping the substance of this evening’s class at the Minneapolis College of Art and Design.
Gartner | Thompson | WaitButWhy | Vinge
This time the speed is the story
As we discussed last week, the history of advertising enjoyed two long periods of little to no technical change. One can argue the stability of those eras enabled great leaps in philosophy and craft. They also bred a kind of comfortable status quo.
Since the Internet arrived, the clock has been ticking faster…and faster.
Technology breeds technology. And here we are entering a second pivotal moment within 30 years. No wonder the pace feels dizzying. That’s a feature, not a bug, however.
The benefit of AI-generative tech is as much about speeding up work as it is deconstructing and enabling all kinds of work. Business analysis, audience insights, media strategy—delivered almost as fast as you can write the prompt (and that’s a skill now, btw). Are the results pristine? Of course not! But they’re worth evaluating, and you got them faster than any previous methodology.
If you haven’t asked Bing’s AI for a competitive S.W.O.T., know an intern at your competition already has. Wharton’s
has demonstrated diverse and surprising ways marketers and advertising people can benefit. But will you try?
Broad impact is also the story
Supply chains, pricing, inventory management, loyalty programs, customer service, and capabilities training will all benefit. Like the arrival of software, then the Internet, AI-generative tech knows no boundaries. (Mollick has been a reliable source for AI-generative use cases, too.)
In class tonight we discussed how these tools feel like a constant companion, a nonjudgemental research librarian. In a matter of seconds we used Bing’s AI to generate a 7-day vegan meal plan including a shopping list. Then some insights on current U.S. teenage chocolate buying habits. Then an optimal route plan for a four-city drive based on local gas prices. Then the raw material to strategize a comms plan targeting moms with 2+ pre-teens living in urban markets. More to the point, these quick exercises provoked enthusiasm for what might otherwise appear to be humdrum tasks.
Our problem idolizing “intelligence”
We’ve spent decades dreaming of this future. That’s kind of the problem. Now that it seems to have arrived, our cultural references don’t serve us well.
To begin with, we humans infer an awful lot—that isn’t there. All the hullabaloo over Bing’s hallucinations says more about us than the pattern matching language model. Conversational tropes like “Hello,” “you’re welcome,” and “sounds fun” are not evidence of thought or thinking. They’re proof a coder built a list of phrases and the model applies them as coded. And yet I catch myself typing “please” forgetting Bing isn't sentient. 🤦🏼♂️🤖
AI-generative tech are wildly potent tools. Not beings.
That confusion is partly why we give them undue creative and/or emotional credit. But the real story is what AI-generative tech is enabling in all of us.
Unlocking creativity, one uncertain human at a time. Or, “Can you draw a fish riding a bicycle?”
Years ago an art director taught me a lesson. “Draw a fish riding a bicycle.” I did, reluctantly. “Look,” he said, “I can tell it’s a fish riding a bicycle!” (The point of the lesson was legibility. Everyone can write words. My partner wanted me to have greater ability to communicate with images. He said you don’t need to be a great artist, just someone who can draw legibly enough for your another person to recognize it.)
Ask your friends and colleagues to draw something, anything, and present it. Even a fish riding a bicycle. So many people will demure, often spectacularly. “Oh, I’m just not creative” is the typical refrain.
Those words are learned, and have never been true. Yet we believe. Thanks to AI-generative tech, “I’m just not creative” might cease to exist.
Because the prompt is not going to replace the pen or the paintbrush, instead it will join them in the creativity toolbox.
Now, ask your friends and colleagues to type the words: “A fish riding a bicycle.” They will look at you weird. But persist. That’s all it takes today to get any of the AI-generative tools to render a legible-enough image of a fish riding a bicycle.
You can draw an image with a pen and ink and paper, if you want.
Or you can type words and render an image, if you want.
The methodology is going to stop mattering for many purposes.
There’s a giant crack now in the Great Wall of Art
I’m pretty sure it was in Randall Rothenberg’s advertising epic Where the Suckers Moon. The history of Madison Ave and Mad People plots a strict organizational divide in the early days: Those Who Write, and Those Who Draw. And Account People.
Those Who Draw were typically Italian. In the days before commercial photography Those Who Draw had irreplaceable talent—they didn’t just conceive the advertising, they created its final artwork, and its typography.
Long story short, the ability to draw, to (ultimately) direct the art, was a rare skill. Pretty much anyone could write. Anyone could have a go at a headline. But not just anyone could Art.
Then Steve Jobs released the Mac. And Joe Pytka directed my favorite Mac commercial… two executives are watching office workers engage with various computers. Exec 1: “Which computer is the most powerful?” Exec 2: “I don’t know, the one with the most mips, or bips, or whatever.” Exec 1: “I don’t think so. I think the most powerful computer… is the one people actually want to use.” (We see a crowd around the lone Mac computer.) Exec 2: “Well that’s not fair. People like using the Mac.”
Then it was a pretty swift trip from desktop publishing to an Internet of images, to Adobe Rush on your iPhone to Canva/Figma. Still, most people couldn’t or wouldn’t communicate visually.
Until now. Until all it takes to make art are words.
Hold on. What about craft?
Let’s agree on a distinction: Sometimes we create images just to communicate an idea—a cue for a partner, a scribble/prompt to gain consensus. That’s where I see the biggest payoff for AI-generative tech. More people being enabled to express visual thinking.
Final, production-worthy art generated by AI? It will depend on the team, won’t it? The thing is, we’re already using all kinds of AI-powered craft in our cameras, in Adobe Photoshop (Sky Replacement or Neural Filters anyone? 🤯), and throughout the video production pipeline.
Wait, what about originality and isn’t AI-generation just stealing?
Seth Godin said it best:
“If an art student studies all of Picasso’s 10,000 paintings and then creates a new painting that is clearly based on them, we call this the advancement of culture…
That’s how culture evolves. Taking an idea isn’t theft. Taking an idea is an oxymoron. Ideas belong to all of us.
We couldn’t and wouldn’t have it any other way. There’s no way to bake a cake, drive a car or write a sentence without using what came before.
GPT and other AI tools don’t actually know anything. They’re pattern matchers and pattern extenders. And those patterns are called culture.”
And we’ll give Seth the final word, too. For anyone concerned AI-generative tech reduces originality, or ideas, or artistry—the opposite holds true.
“If your work isn’t more useful or insightful or urgent than GPT can create in 12 seconds, don’t interrupt people with it. Technology begins by making old work easier, but then it requires that new work be better.”
I’m excited to see how much better our creativity will become.
Don’t forget the survey!
I’ll have an “After” reflection in your inbox late Wednesday.
If you’ve read this far, answered the survey, and aren’t yet a subscriber, this button is meant for you.