I don’t know if this will land the way I want it to, but I’ve been having versions of this conversation with a lot of creative people I know - musicians, DJs, UI/UX designers, architecture students, people in fashion - and it keeps pulling me back, so I’m writing it down. I exist in this weird in-between place: I’m technical, I work in AI, but I see the world through the lens of art - particularly hip hop music. I released an EP at 14. Mr. Morale & The Big Steppers changed my life in ways technology can only dream of. And I’ve spent years feeling like I didn’t quite belong to either world because of it.

And this is the thought I keep having working with AI agents: if the new interface is just your language, the one you’ve always used to say exactly what something should feel like — then creatives might be exactly who this moment was made for.

I think that’s why I want to write this.

If you’re a creative, you’ve probably spent years building up a sensitivity to what feels right, and I mean actually right, not almost right — the exact pacing of something, the exact texture of it, the difference between something that works and something that genuinely lands, and you’ve built your whole process around the tools that let you get there, the software you know, the workflows that feel natural, the things you can personally execute — but there were always ideas sitting just outside that boundary, the thing that needed a developer friend, the vision that required a skillset you didn’t have and weren’t sure you wanted, and so it just sat there.

That gap is getting smaller.

Earlier this week me and some friends went looking for Pokémon cards, specifically the new Ninja Spinners pack, went to a couple of places and a few seven-elevens and couldn’t find it anywhere, sold out everywhere we checked. We’d seen our friend Tar open packs before and it looked genuinely fun, that whole dopamine hit of not knowing what you’re gonna get, and we wanted that and came up completely empty.

That evening I had this kind of dumb idea. What if I just built a simulator? A pack opening thing online, for that specific pack, with the actual cards from it. And I asked the AI — and I want to be honest about how unspecific the prompt was because I basically said something like: I can’t find Pokémon cards nearby, can you just build me something that scratches that itch?

That was literally it.

It found the official Pokémon website, which apparently leaves its entire card database completely open — every card, every image, every pack — and it built the simulator from that, animated the cards flipping, matched the (estimated) real-world odds for rare pulls, and I didn’t ask for any of that specifically, it just understood what a satisfying pack opening experience would actually need to feel like and built toward it.

Later that evening I’m at HVEN in Nakameguro, a venue where my friend Max DJs and they do open deck Wednesdays where it’s pretty casual and anyone can come hang, and I showed him this thing on my laptop. Max is not a coder. He just has taste. He looked at it for maybe thirty seconds and said: yo, make it like Balatro, make it addictive, MORE flair, MORE dopamine, just more of everything.

So we started telling the AI that. More. Make it more addictive. And it started adding things we never mentioned — the phone vibrates when you pull a rare card, there’s a combo system, if you open cards fast enough, under 500 milliseconds, you get a multiplier, there are sound effects, there are mechanics we genuinely never thought of — all because it understood what “more addictive” actually meant and just ran with it.

Max and I were sitting at a bar on a Wednesday night saying “more” and an AI was building a game in real time.

And then I thought — I already have a domain, I already host this blog on a Raspberry Pi (its a small computer) at home, can I just put this on there? Told the AI that and five minutes later cardgame.cruijffiaan.ink was live and anyone in the world could play it.

From dumb idea to playable website, the whole thing took less than an hour, and honestly most of that hour wasn’t even really building — it was just reacting to what was there, figuring out what felt better, and pushing it further.

The interface to all of this is natural language — your language, the one you already speak — and I think that sounds obvious when you say it out loud but it actually isn’t, because everything up until now required you to speak a completely different language first, to learn the software, to understand the tool, to acquire some skill before you could even begin to express what you were trying to make. And now you just don’t. You say what you want and it happens, and the more specifically you can describe what you want, the better it gets.

I have a friend, Taymoor — he’s creative in a way that’s hard to describe, it’s not “I’m a musician” or “I’m a designer,” it’s more diffuse than that, it’s taste, it’s the way his mind moves, it’s that he has an instinct for what feels right across a lot of different things. He’s built websites before, but a terminal screen was new territory, so when I showed him Claude Code last September he looked at me like I was handing him something in a language he couldn’t read. “Okay, what is this. What can I do with this.”

He was quiet as I showed him around for maybe five or ten minutes. And then he started asking questions - not confused questions, pointed ones. “Wait, I can create a website from here? I can do this from here? I can do that from here? What about this - I’ve been wanting to do this for the longest time?” He wasn’t asking me what the tool could do. He was running through a list — his list, the one he’d been carrying around — checking each item against this new thing he was holding, seeing if it cleared. That’s what an idea bank looks like when it opens. Not a flood. A series of very specific withdrawals.

Within a week he’d had so many ideas (one of which was “we should livestream us Claude Coding while kicking back a couple of beers”) and applied it to so many parts of his life that I lost track of what he was working on. But there was something else I couldn’t stop thinking about - I had known what this tool could do before I showed him, and watching him receive it was still nothing like I’d ever felt before.

Technical people could always will things into existence. That was genuinely never the constraint — we had the tools and the skill and if we wanted to build something we could. What we didn’t have, and honestly what most of us still don’t have, is what you have. The vision. The instinct for what makes something feel right versus almost right, the ability to walk into a creative process and just know. The thing that’s shifting right now isn’t who matters more — it’s that the constraint has moved. It moved from execution to vision. And creatives have been training for exactly this their whole lives, without knowing that’s what they were doing.

Because if the cost of making drops, then the value of knowing what to make rises.

And that means the people who have spent years developing taste, instinct, and sensitivity to what something should feel like may be far more central to this moment than anyone realizes.