Play freely. Control precisely.

A designer’s case for building creative tools around play, instant control, and automation working together.

Why I built Whorl, a new design tool

I’ve spent much of my career as a designer, building tools for creativity, designing both what they let you make and how you make it. The way I’ve typically done that is by playing: prototyping, trying things, following what felt right until a new kind of interaction emerged that I wouldn’t have specified in advance. Enough of those turned into real, novel things that often generated patents. Play, it turns out, wasn’t the opposite of the serious work. It was how the work got done well.

One random evening I was dabbling with a bit of code to generate visuals in motion, watching shapes move on a screen in ways I hadn’t quite planned. That feeling was play too. And it turned out to be the experience I most wanted to give other people.

Whorl started as a small iOS app I made after my time at Adobe, sort of as a way of bottling that feeling. It found a quiet following among casual creators, and I mostly left it alone for a few years. To my surprise, it kept being used, kept finding new people on its own. And they asked for more. There was a bigger, more capable tool hiding inside that little app, and I wanted to build it: something that could hand you a whole result in a click and still let you shape every detail by hand. So for the last two years I have been turning Whorl into a proper tool for people who make things for a living. I designed and built the experience, with paid engineering help on some of the harder technical foundations, backend, data, and render-engine performance. Through all of it I held onto the one thing that made the original worth anything: it is fun to use. Only now it is powerful enough for real work as well.

Who it’s for, and what you make with it

Whorl is a generative design tool for making visuals, either still or in motion. Posters for an event or announcement, brand and art assets, backgrounds for a talk or a presentation, loops for a music video, visuals for screens on a stage like conferences and VJ sets. The output can be a calm still or an expressive piece with motion that holds a room.

At the core of it are the stencils: sets of generative style stencils, each one carefully designed as a family with its own range of possibilities. One stencil might give you bold forms, another fine textures, another the kind of layered depth that reads as a 3D optical illusion, another soft, stylized backgrounds. Each one holds endless variation inside it, so a single stencil is not just one look but a whole territory to explore.

Whorl is both goal- and process-focused. The process is play, but the goal is to arrive at something you actually needed to make. So Whorl brings together the things you’d usually have to stitch across several tools: the abstract visual itself, text and layout, and motion, all in one place. You can start from an abstract form and end with a finished poster, titled and composed and moving, without leaving the tool or exporting halfway through to assemble it somewhere else.

That combination is deliberate. An abstract visual can be a finished piece on its own, a vibe, a mood, open to interpretation. But when you need it to communicate something specific, an event, a post, a message, you need text and layout too, and Whorl composes them together with the visual. Add motion, and you have a tool for making complete, unique communication. The layout does real work for you, holding the composition together instead of leaving you to arrange every element by hand, but more on that later.

Whorl is built for anyone who needs to make something striking: designers, art directors, motion and visual artists, and VJs, but also event organizers, marketers, and anyone putting together a talk, a poster, or a moment on a screen. It is playful enough to pick up the first time with no design background, and deep enough that a professional can make something they’d put their name on.

The balance I keep chasing

Here is the design problem I find myself returning to, over and over. It is genuinely hard to make something that produces complex, striking work, gives you enough control to make it your own, and still stays intuitive and playful to use. Those goals fight each other easily.

Powerful tools tend to get heavy. They accumulate functionalities, UI panels, modes, and controls, a learning curve you have to climb before you can make anything good. Fun tools tend to stay shallow. They are easy to pick up precisely because they can’t do very much. The moment you make a simple tool more capable, it usually stops being simple. The moment you make a deep tool more approachable, it usually loses its depth.

My work in general, and Whorl in particular, has been an ongoing attempt to refuse that trade-off. We need more tools where you can make something striking in seconds, or lose an hour in the same session going deeper, and have both feel good. Where randomization could hand you a direction you would never have drawn by hand, at least not quickly, and then detailed controls could let you shape it until it was unmistakably yours. Surprise and control, moving between them however you like.

Why play

I keep using the word play, and I mean something specific by it.

When you play, you try things without knowing where they lead. You follow a curiosity instead of a plan. You make a mess, keep the good part, and go again. You discover and learn by doing. Play is how children learn, how musicians find a riff, how a lot of the best design actually happens before it gets cleaned up and rationalized after the fact.

Most creative software is built for the cleaning-up part. It assumes you already know what you want and just need to execute it efficiently. That is useful. But it skips the part I find most valuable, the part where you discover what you want by messing around. Whorl is built for that part. You push a control and the whole composition responds, instantly, and that instant response is what invites you to interact again, and again, until you have wandered somewhere you couldn’t have planned your way to.

Sometimes I start with an idea, and play is how I chase it down, testing whether it holds up, finding the version of it that actually works. Other times there’s no idea at all. I’m just playing, and something appears that I wasn’t looking for, and that becomes the idea, or the start of something I hadn’t imagined. Both directions are real, and in practice I think people work this way far more than they let on. We’re taught to have a plan, a process, a defined path to a result. But a lot of the good stuff arrives sideways, by playing your way into it. A tool built for play makes room for both: the idea you came in with, and the one you didn’t.

Real-time isn’t a technical detail here, it’s a foundational user experience. Direct manipulation, where you act on something and it responds instantly, was there at the early days of interaction design. In 1968, Douglas Engelbart’s Mother of All Demos showed live, operational tools responding in real time, long before personal computer operating systems as we know them today. Sometimes new technologies make us drift away from this. But machines are extraordinarily fast now, even in a browser, and I think there’s still a lot of room to build on what that 1968 demo pointed at. The loop of push, see, respond, push again only works well if the response is immediate. Any delay, any pause where you have to wait and evaluate, breaks the flow, making it feel like work instead of play.

Designing for play

Play needs a low barrier to entry. When we play, we don’t think much about how or why. Ideally, we don’t prepare, we don’t read a manual, we just start, motivated to try things and see what happens. A tool that makes you stop and figure it out has already broken the spell. So a great deal of the work in Whorl is invisible: it goes into making the tool subordinate to the creation, so it disappears and the making can begin instantly.

It starts with the overall structure. There’s nothing innovative here, and that’s deliberate. The layout is clear and conventional, so you can take it in at a glance and know, spatially, where things live and where to go to navigate or interact. You don’t have to learn the map before you can start. That high-level clarity is what lets you stop thinking about the tool and start playing with it.

Then there are the many small things. The size and arrangement of the controls, how they’re named, how they respond, how edge cases are handled so nothing ever feels broken. Standard industry components aren’t always the best they can be, so I build custom or adjusted controls to refine how they behave. Most web-based slider draggers, for instance, jump to your cursor the instant you grab them, throwing away your starting point right when you’re trying to make a small, precise adjustment. A little disruption, but the kind that bothers me, so I built my own. The interface also adjusts to match whatever you open: load a starter or one of your own files and the panels focus on where that piece is set, so you land in context and can pick up where you left off. On their own these might seem trivial, not critical. Together, they are the difference between a tool that feels effortless and one that quietly fights you.

But ease isn’t only about removing friction. Sometimes it means inventing a way to do something hard that no tool does well. Color is a good example. In Whorl you can pick or create a color palette and watch everything on the page, the background, the text, and the Whorl itself, shift together in a harmonious way, with the text staying legible throughout. That legibility part just works. Add different color distribution methods, and you can see your entire composition in a whole new color theme in a wide range of possibilities in a single click. In most tools this is slow, manual work, and where it’s automated at all it tends to be too basic to trust, or it falls apart across the whole composition. Getting it to work beautifully and holistically took a long time. It is one of the things Whorl does exceptionally well.

And this is where the craft loops back to play. The faster you can try something, the more adventurous you become. When changing an entire color theme costs one click instead of an hour, you try the version you wouldn’t have bothered with, and that’s often the one you would never have arrived at any other way. Ease isn’t just comfort. It is what makes discovery cheap enough to be generous with.

The same ease makes variations almost free. Because randomization runs at every level, you can make not just one poster but a coherent set of them, shifting the stencil, the color, or the composition across the series while keeping them clearly related. A whole family of outputs that plays off itself. That’s normally slow, careful work, and here it’s fast enough to stay playful, so you make the series instead of settling for the single piece.

Layout works the same way. There are five layouts, each with a range of variation and built-in rules for how the elements relate, so Whorl holds the composition together as you go. Change a line of text and it repositions the rest for you, instead of leaving you to nudge every element by hand the way most tools do. Small options build on this. You can inset the whorl into a rectangle on the page, for instance, which changes the relationship between the art and the text, and everything else repositions itself around it automatically. It’s a small feature with barely any interface, and it opens up a whole range of compositions you’d otherwise have to build by hand. That’s labor you’d rather not spend, freed up for trying things instead.

Motion works on the same principle. In most tools, animation means a timeline and keyframes: you place them by hand, set values, nudge them around, fuss with easing curves. It’s powerful, but it’s tedious and technical, the opposite of play. In Whorl you reach for motion presets instead, applied instantly, with detailed controls underneath when you want to shape the timing and feel exactly. It’s built around loops for now, seamless and continuous. The method works, and I’ll expand on it over time.

The AI consideration

People sometimes ask whether there’s AI in Whorl. There isn’t, at least not for the person using it, and that’s worth explaining, because it wasn’t a reaction to anything.

I’m not against AI. I worked with machine learning long before the current AI wave, and I’m drawn to any solid or emerging technology that on its own or as a foundation can make for a better experience. AI is also a bigger and more layered subject than Whorl’s design choices, one I recognize doesn’t reduce to a single opinion. But my decision here was narrower and more practical than any of that. When I started building Whorl, coding with AI was still basic. I used it, and it helped a little. Over time it became a real enabler, and I leaned on it more and more, mostly to build the technical foundations that let me create the experience I was after. The experience itself I shaped by hand, with meticulous control over how every part looks and behaves. AI helped me move faster, and even reach ambitious ideas I’d assumed were for much later. But using AI to generate Whorl-like results never fit the playful spirit I’m after.

Not as a stance. It came down to one thing: waiting. A prompt is a request you send and then wait on, and that wait is exactly the pause that breaks play. Play depends on instant results. You push something and it responds right away, and the immediacy is what invites you to push again. The moment you are describing what you want to a text box and waiting for a result to come back, you are in a different, more transactional kind of playground. For a tool whose whole value is immediate, hands-on discovery, that wasn’t a step forward.

So I left it out. What I kept instead was a foundation of curated generative results, real-time response, and controls you can feel your way through. That turned out to be more powerful, for this particular tool, than anything a prompt box could have added.

The best of both worlds

I don’t believe the future of creative tools is either fully manual or fully automated. It’s both, working together, on the same material.

Whorl is full of automation. You can start or randomize at every level, the file, the stencil, the shape, the color, the motion, the layout, the text. An algorithm hands you a whole composition, or reharmonizes the color, or moves the elements on the page to keep them composed. There’s a lot Whorl’s logic can do for you. Not because doing the work by hand doesn’t matter, it often is the making, but because you shouldn’t have to do all of it, all the time. Automation is there when you want speed, or a starting point you wouldn’t have thought of. Manual control is there when you want to shape something exactly. You choose, moment to moment, and the tool keeps up either way.

But the automation and the control aren’t separate halves. They act on the same thing. Whatever the logic proposes, you can take hold of and adjust by hand, the properties of the exact thing it just generated, and because Whorl is procedural, nothing is ever locked. You can override, revert, and try again, non-destructively, in real time. Whorl’s logic proposes, you steer, and neither one takes the making away from the other.

That’s the balance I care about most. Not automation instead of you, and not everything by hand, but a well-crafted algorithm and a well-crafted interface meeting on the same canvas, so the tedious parts get easier and the creative parts get bigger. Automation that gives you more room to play, rather than less reason to.

Where it’s going

Whorl isn’t finished. It’s the opposite, really. Every capability it has opens up three more I want to build.

The direction I care about is expanding what you can make, especially the things that are hard to create anywhere else. Not more of the same, but new kinds of visual form and motion you haven’t considered, or were too intimidated to try to create. Some of that lives in dimensionality, giving the flat forms real depth, in look and in motion, letting them move through space. Some of it lives deeper in the generative and procedural side, the kind of effects that can turn a simple starting point into something rich and surprising. The common thread is unique output: forms and motion with a quality that’s difficult to fake, that reward playing your way into them. And whatever gets added stays as playful to reach as everything else, because the how matters as much as the what.

I’m deliberately not making a feature list here. Part of building this way is that the best directions tend to reveal themselves through the same process the tool is built on, by playing, trying things, and following what surprises me and the needs I hear from the community. What I can say is that the goal never changes: keep making complex, striking work easy and playful to reach, and keep finding the things worth making that almost no one else can.

Play freely. Control precisely. Both on the same canvas.
Come play at whorl.app

Play freely. Control precisely. was originally published in Bootcamp on Medium, where people are continuing the conversation by highlighting and responding to this story.