The AI backlash won't save us
Welcome to Fort Human, a Descript newsletter.
By Brandon Copple
A couple months ago Swen Vincke, CEO of the video-game maker Larian Studios, told a Bloomberg reporter his team was using AI to generate concept art and create PowerPoints to present those concepts. The comment is buried two-thirds of the way through the resulting article; you could easily miss it.
Larian fans didn’t miss it, and they didn’t like it. The backlash was swift and severe.
Vincke got a social-media whipping for daring to use AI in game development at all; Larian fans chastised him for replacing concept artists with AI even though he didn’t say that; they went upside his head with the slippery-slope—first you came for the concept artists. Many promised never to play Larian’s games again.
Poor Swen had to issue a statement promising he was not replacing any artists with AI. And he wrote an amusingly defensive tweet.
But the comments crowd, by now drunk on their rage, only lashed out harder.
This kind of backlash can be helpful for us creatives. If it causes the folks leading creative teams to think carefully before they run headlong into generative AI, it will buy us a little more time to figure out which end is up.
But the anti-AI crowd isn’t going to protect everybody.
Crickets for Rocket
A few days after the Larian thing blew up, Rocket Mortgage released a new commercial its creative team made for the NFL playoffs. In interviews, Rocket’s executive creative director said they’d been able to produce the ad entirely in-house—without an agency’s help—partly with the help of AI. AI tools helped them land on the idea faster, smooth out feedback (or speed it up, or both), and get to a good result without outside help, he said.
That article is light on details (there are none), but it sounds to me like the Rocket team was using AI in a similar way as Larian. The leaders of both teams say they were just looking for ways to make the creative process work better. But whereas Larian’s boss swore up and down wouldn’t use AI to actually create anything or replace anyone, Rocket’s leadership full-on bragged that they were going to “restructure the whole team around artificial intelligence”—and cut their ad agency out of the picture.
Ad agencies, as you may know, consist primarily of humans, including many creatives—most of whom wear bulky, stylish glasses and all of whom would prefer that their clients didn’t replace them with AI.
But that’s what Rocket did. And yet, while the internet was beating Swen Vincke’s brains out, as far as I can tell nobody said a word about Rocket.
We’re on our own, kids
What I think these two stories, taken together, tell us: if you’re a working creative, producing stuff for a business, your audience isn’t going to save you from AI the way they might if you were creating for media or the arts.
Games, movies, music, books, magazines, even newsletters and video channels, create visceral connections with audiences, even a feeling of ownership.
People just don’t feel that way about business content. So there won’t be any angry crowds of commenters rushing to our defense as AI seeps into our workflows.
Yes, there might be some pushback if your company fired the entire creative team and started churning out slop. And yes, some brands have taken a whipping for releasing AI-generated ads — but they keep doing it, like Svedka just did, at the Super Bowl, to pretty muted response.
Anyway if you ask me all the attention on AI-generated ads is a distraction. For the vast majority of creative teams, the big changes won’t have anything to do with expensive, splashy commercials or even generated video.
Most of the changes in the coming months will be on the back end: using AI to research, brainstorm, story board, create early drafts and prototypes, write creative briefs. The kind of stuff that Larian and Rocket are doing — that sent Larian’s customers into a rage, but not Rocket’s.
Those of us who earn a living doing creative work for a business are far more likely to work at a Rocket than a Larian. We’re right there in the boat with Rocket’s creative team and their agency (which, happily, did work on the company’s Super Bowl ad).
This is why I think it’s imperative, and urgent! and critical! that we start figuring this out for ourselves. If creatives don’t find the right ways to use AI in our work, somebody else will do it for us. And their definition of what’s right will be different from ours.
This is a fundamental belief here at Fort Human. We didn’t ask for AI, but it’s here. Love it or hate it, it’s going to be part of creative work. So we creatives have an obligation, to ourselves if nothing else, to try and use it to make our work better, do it faster, and maybe (the tricky part IMO) make it easier on ourselves. We can do that by trying stuff, and by educating ourselves on how AI works, what it does well, and what it doesn’t. We should be clear-eyed about the risks and downsides — but also about the ways AI can be useful.
I can’t say for sure, but I’d bet that most creative people agree with most of that. If you do, I hope you’ll keep reading.
Wait, what did I just read?
This is the first newsletter from Fort Human, a weekly newsletter by Descript. It’s going to start showing up in your inbox, usually on Wednesdays. If you’ve ever read the Descript newsletter, you sort of know what to expect.
Here we’ll explore the shifting landscape of creative work as it’s transformed by AI. So there’s your “fort” metaphor.
You’re receiving our inaugural post because you subscribed to the Descript newsletter — which is now Fort Human.
You can read more about Fort Human, our editorial plan, and our perspective on AI, on our About page. We hope you’ll like it. Let us know either way, in the comments here or anywhere you follow Descript.
What fresh hell
AI news for creatives, as summarized by the Claude chatbot, given this prompt:
“You are an astronaut who’s obsessed with The Voice; you just found out you got tickets to a live taping, but you can’t go because you are in space.”
Amazon is reportedly planning a content marketplace where publishers can license content to AI, and a novelist claims she used AI to write 200 romance novels last year; if she wants a break maybe she can take my Voice tickets, since I’ll be busy hurtling toward the ISS at 17,000 miles per hour. People weren't sure which Super Bowl ads were AI generated — try watching it through a porthole, on a 30-minute delay, which is how I’ll be watching the knockout rounds — and New York is considering a law requiring labels on AI-generated news. I’ll be up here manually recalibrating oxygen scrubbers instead of sitting in the room when John Legend tells somebody they have “it” but at least the new Seedance 2.0 can generate entire short videos with sound effects. Meanwhile, AI chatbots are proving adept at diagnosing medical issues, causing some doctors to wonder what doctors are even for now—welcome to the club, doc.
Specifically re: Larian getting tomatoes thrown at them over mild, 'backend' AI use and not ad agencies, Rocket Mortgage, etc. --
I agree that 1) people care viscerally about disruption to the arts and not to commercial content but also want to add 2) people yell at who they think will listen. Brands that people affiliate with their own identities, whether game-makers, indie streaming platforms, fashion houses, etc. are perceived as 'extra accountable', reactive, likely to respond to online invectives.
It's why people pressure seemingly random tiny arthouse presses to come out with special anti-AI statements.
The biggest problem this ongoing furor causes is not, you know, all the anti-AI cancel mobs...
but that if the prestigious creative class is against generative XYZ medium, the best aspiring creatives in that medium will also avoid AI tools. If you're trying to make a tool or a platform for AI games, AI movies, AI music, whatever, you're going to instead systematically attract casual dilettantes and find it much much harder to build any real creative ecosystem.
The medium is more important than the method... at least with video games vs. home mortgages it seems.
Based on backlash like this, why would leaders be honest about their use of Gen AI in the future?