Creation becomes consumption
What if the next generation of creative tools felt more like authored experiences than blank canvases?
It’s 2032. My favorite movie this year? I generated it. Sean Baker, one of my favorite directors, built the tool. I didn’t study cinematography. I didn’t touch a timeline. I opened an app, made a few choices, and hit play.
The illusion of authorship
Today, I type unfinished thoughts into a box, drop a few references, and hit enter. The model handles the rest. I get a clean result that feels like something I made. But mostly, I curated inputs.
During Ghibli Day, it felt like we were making. We fed photos into an AI model trained on the world of Miyazaki. It was fun, multiplayer and internet magic. But was it creative?
“They gave us an everything generator, but everyone is obsessed with making it do the same thing.” — Luke Miles, on Ghiblification
Our role was reduced to prompting and curating. I dragged a photo into ChatGPT and generated three versions. Picked the one I liked, sent it to the group chat with an unspoken: look what I made.
But did I really make anything? Or did I just consume via prompts? That distinction - between creation and consumption used to feel obvious. Am I filming or watching a movie? Now it doesn’t.
As generative tools become faster, more aesthetic, more automatic, “creation” starts to feel like consumption. Creative tools stop behaving like neutral platforms and start acting more like media, like watching YouTube or playing a game. So what happens if we design tools that lean into that? Not generic platforms with infinite flexibility — but tools that are expressive, opinionated, guided?
Neutrality is a myth
Most creative tools today try to be neutral. They try to be for everyone, leading with clean interfaces, neutral defaults, and infinite possibility. Figma doesn’t tell you what to design. Final Cut doesn’t suggest how to cut. Loveable lets you build anything.
But with generative AI neutrality is a myth. The defaults shape the work. The training data always shows. I can tell it's a Lovable site - same iconography, same yellow-tinted generated image. Same with Midjourney. The tools and models leave a fingerprint.
So what if we stopped pretending the tool was blank?
Tools with taste
Instead of hiding the taste, what if we made it visible? What if tools had style? What if they guided you toward an aesthetic — not just in the output, but in the process itself? Not templates onto generic software like Squarespace or Webflow, but taste baked deep into the workflow: the colors, the rhythm, the worldview.
In that context, constraints become generative.
They don’t limit creativity; they shape it.
Tools as media
When creation becomes consumption, tools stop behaving like tools and start acting like media.
Most people don’t want a blank canvas, they want a dance, a format, a prompt. If the tool doesn’t offer that structure, the feed will (“make it Ghibli”). When everything is possible, the strongest constraint wins.
Toolmaker as artist
At Danger Testing, we’ve started designing tools with a point of view.
Our collage app uses fixed shapes, a locked palette, and the ambient sound of Japanese train stations. You can’t go off-grid. But follow the rhythm, and you’ll land somewhere cool.
Curate Curator is another experiment. You build a persona. The system gives you four cultural recommendations: one song, one book, one painting, one poem — all on fixed, pre-colored backdrops. Rick Rubin offers guidance and the system prompt references “032c”. You’re not generating “anything.” You’re stepping into an authored lane.
Expressive constraints
I used to think functional and expressive software were separate categories. One clean, one chaotic. One Dieter Rams, one Michel Majerus.
Now I’m not so sure.
I think the most interesting creative tools going forward won’t try to be everything. They’ll choose a lane, express a point of view, and build an experience around it. In this world, the tool becomes a collaborator with strong opinions about what good work looks like. You don’t choose tools purely for functionality. You choose them like you’d follow an artist.
Hans Ulrich Obrist framed curation as “creating a structure — a score — in which others can perform.” His do it exhibition didn’t show finished works. He asked artists to write instructions anyone could follow. The art wasn’t the object. It was the act.
That’s what tools are becoming. Not utilities, but collaborators. Systems with taste. Structures with rhythm. Places you step into, not just software you use.
The process becomes the product.
The constraint becomes the style.
The toolmaker becomes the artist.
Maybe the next Casey Neistat isn’t a filmmaker.
Maybe they build the experience you film inside.
Maybe they design a story that writes itself with you.
Thanks to Carly for the edit!!
 | A guest post by
|
Beautiful piece, Marc! I think we’re already seeing some of this in hardware as people trade some experiences on their iPhone for more dedicated devices like e-readers, iPods, gaming systems, etc.
Even though everything can be made for the iPhone, maybe it shouldn’t be an everything device? The same can be said for software, and you’ve illustrated that beautifully.
Thought-provoking words, Marc. Thank you for this!
I agree with your take on creation blurring into consumption, and the idea of curation as a kind of authored performance. Also, I think we need to bring more intentionality to how we frame synthetic outputs as legitimate media. If it meets the formal requirements of the medium, is that enough? If it looks like a movie, is it one? We may need to overhaul our artistic lexicon more radically, especially as some creators continue to forgo AI-powered tools (for as long as possible), thereby creating different tribes of creatives.
That said, maybe the answer isn’t to discredit synthetic output either, but to acknowledge it as something distinct, with its own criteria. Think ‘Best Synthetic Film’ at the Oscars. That could help us preserve room for different types of authorship, rather than collapsing them into one.