Uncomfortably numb: making people feel in the age of AI
When you use AI to push creative boundaries instead of treating it like a shortcut, you don’t just make work faster, you make different work possible. We’ve seen how this plays out: new technology doesn’t automatically create better work, it raises the stakes on taste.
When Terminator 2 landed, it created a cultural shock. The first time the audience saw the villain sliding through prison bars, it didn’t just impress; it became a new benchmark in how technology can impact culture. The audience embraced the movie and the choice of using CGI, because it allowed us to tell a story we could not have told otherwise. The thrill of liquid metal as a villain - it served the storytelling. And everything else from the previous film stayed the same. The makeup, the glowing eye, the accent.
A few years later, George Lucas took a different approach with his Star Wars prequels. The technology didn’t open a new door, it moved into the center of the room. Practical texture was traded for digital smoothness. The puppet became pixels. The lived-in world became a showroom. It lost its grit and humanity. It looked “better,” but it felt like less. And the audience felt betrayed. Because what they loved wasn’t the franchise’s ability to look futuristic, it was its ability to feel real.
The AI Moment
Generative AI has reached mass adoption faster than any previous technology shift, including the internet, but what’s becoming clear is that speed of adoption doesn’t equal depth of transformation.
It is everywhere now. But what it is actually changing is still an open question.
And so, next following question would be, are you using AI to solve the one problem you couldn’t solve without it, while you keep doing the hard, human, messy work everywhere else? Or are you using it to skip the craft, remove the friction, and quietly sand down the very imperfections that make people care?
We’ve already seen what happens when brands use AI as a shortcut instead of an elevator. Coca‑Cola’s AI-generated holiday homage sparked backlash for looking “soulless” and “uncanny,” with critics reading it less as innovation than as swapping craft for cost-cutting. Toys “R” Us premiered an AI-made Cannes film using OpenAI’s Sora and immediately got hit with the same charge: that it was “AI slop” trying to sell imagination while quietly removing the humans who make it.
And beyond advertising, the line gets even sharper: an “AI George Carlin” comedy special triggered outrage and a lawsuit from Carlin’s estate, explicitly framed as needing to “draw a line in the sand” against unlicensed imitation. In every case, the audience reaction wasn’t “don’t use the tool.” It was: don’t make the tool the point.
Elevator, not an escalator
So here’s the mandate: use AI as an elevator, not an escalator to the same old floor.
If the brief is constrained by budget or time, don’t use AI to deliver the same output cheaper. Use it to deliver something truer, stranger, more ambitious, something you couldn’t have shipped before. Let it expand the palette, not flatten the process.
Because people don’t care how the work was made. They care whether it makes them feel something. According to Ipsos, Americans’ descriptors for AI-generated content have rapidly flipped from “futuristic/innovative/creative” in 2023 to “fake/soulless/not ‘real art’” today. That isn’t a rejection of innovation, it’s a rejection of work that feels engineered around what the model can do, instead of what the story needs.
Treat AI like a creative constraint. Not a replacement for ideas, but a pressure test for them. Not a way around taste, but a demand for more of it.
In the next year, the brands that win won’t be the ones using AI the most, but the ones using it to make work no one else can. Ask the only question that matters: what could we make that couldn’t exist without it? That’s your liquid metal moment.
