Lars Wander

I found my way into computer artwork through my job at Google, where I was working on a team called Visual Analytics. It was a small team with an interesting mission that involved trying to build visual intuition for systems that were too big and too complicated to understand with traditional data-based techniques. At the scale Google operates, there's often just too much data and complexity to wrangle using conventional approaches.

My mentor and manager, Wim De Pauw, had been doing this kind of work for over 20 years. It wasn't until I saw what he was doing that I thought, for the first time, "Wow, you can really make beautiful things with just a computer program." That had never even been a possibility for me before.

What I found so freeing was discovering pen plotting around the same time. A pen plotter is like a 3D printer that can only go in two dimensions: it holds a pen and moves it across paper. While I could leverage the same computational techniques I used at work for rendering shapes and lines, I found much more expressiveness in exploring algorithms and techniques for sake of visualizing their output than the systems I was drawing for my job. It was much more fun to express myself using these tools without requiring the same degree of precision needed for data visualization – the right color schemes, contrast ratios, all those mechanical aspects of the data-visualization side.

Going back to my Google team, the most fun we had was building a visualization for the first time where we had some idea of what we wanted to expose, then drawing it and seeing a completely different world open up. Some behavior the system was expressing that was invisible to us suddenly became apparent, and it was never something we could predict.

What I enjoyed about applying this to computer artwork was thinking about systems where the underlying behavior really couldn't come to life without a visual. We've internalized in computer-based art that simple rules can lead to very emergent and complex behavior. But there's always the challenge of wrangling that complexity into something visually compelling.

It comes down to what you want to say with the art and how you connect it back to the world around us. While it's interesting to take something like cellular automata and make it simulate the universe, its real strength comes from looking at rules around us in the world we live in, and how those give rise to the kinds of complexity we see organically in nature.

The biggest thing I'm realizing is that what's really important is not only uncovering something emergent, but also being able to wrangle it to some degree. If you're completely hands-off and have no idea what your code is doing, it runs away from you very quickly. Sometimes there are fleeting moments where things look interesting and beautiful, but overall it leads to a system you can't really express your ideas through because you don't understand it.

An example was when I wrote my first spectral ray tracer – a system that tries to render light and its path through an environment in a way that's physically closer to how light behaves in reality. This spectral ray tracer takes into account frequency and phase information, forming what's known as a spectral power distribution, which models how light enters your eye and hits your cone cells to be converted into color.

Just by including this extra bit of information – the frequency and phase of light as it travels through the scene – a huge range of complex behaviors opens up. Hold up a light to a prism and you get a rainbow that's far more interesting than light just bending in monochromatic white.

Writing something like this is complicated and takes weeks of work, and you end up with bugs. Those bugs often produce even more interesting behavior than you intended, but it's very important to isolate them, understand what they're doing, and then use them as a creative tool, but not leave them as something you can't explain.

So much of art making is a deliberate set of decisions. At every point when it comes time to decide one thing or another, you're left wondering: is this decision important to the story I'm trying to tell, or is it not important? In cases where it's important, I don't feel it's right to leave it up to chance because I want to craft the narrative in a certain way.

But in other cases where something isn't strictly core to the system or story being told, I think it's wonderful that with computers we can introduce what's almost like true randomness. We can say very deliberately that this is a decision that's maybe more interesting to be left to chance than to my discretion as the artist. Of course, that itself is a decision you're making.

What's interesting with computer-based artwork is that you have this very precise degree of control, and ironically, that precise control is all about letting go of control in those moments.

You're exploring parameter spaces where you have a few inputs, and each input is a dimension in these very high-dimensional spaces we create. They feel infinite, but the really mind-boggling thing is that the number of states our computer can take on, no matter how big, is always going to be inherently finite.

This makes it sound like a smaller possible space than you'd think, but also makes finite numbers far bigger, in my mind. I'm torn both ways – it's "only" finite, but finite means a heck of a lot. Like the finite number of thoughts we can have, even thoughts we could have in our lifetime. That sounds small, but it's a testament to just how big numbers can be.

Absolutely. The classic philosophical question we all go through is "Is my red your red?" You can't answer that question, but what's fascinating is that despite this being inherently unanswerable, almost a hundred years ago we came up with the core foundational models of color we use today to understand everything from printmaking to designing screens.

These models of color are very precise, but what's almost magical is that they link this rigid mathematical definition of color with this very fluid subjective one. With just a little bit of math and some filters, we can predict exactly what color your brain is going to tell you you're seeing. That feels pretty absurd, especially considering how long ago people figured this out with such precision.

These questions about how we see the world, and the fact that with computers we can model many elements of our reality to a very high degree, makes this link between perception and computation interesting. We have tools that can break down these barriers and break down our understanding of reality just by tweaking simple parameters. I find it very gratifying to search and chase those kinds of ideas and projects.

This idea that you can't really untangle observation from interaction feels at first completely alien, but the more you sit with it, it almost feels like an inevitability. I love those kinds of concepts that at first can't be true, but then seem obvious, then feel impossible again – they oscillate back and forth.

If you observe something, is there a way to truly be a passive observer? At first you'd say yes, you can just look at something without touching it. But when you think about what observation is – at least with your eyes, it requires shining light on the object, and that light has to reach the subject and come back to your eye. We're not very attuned to how much light pushes on our skin or how warm it can be, but it's inherently interacting with the system we're observing.

This really grounds me in the world. We're part of the universe, not free-floating minds existing outside of it. We're embedded within it. Any time we see, hear, or feel something, that comes from an interaction with the same matter we're made of.

I definitely feel that ambient anxiety. Anyone with enough computer background sees how quickly these systems evolve and become obsolete. I was recently at an exhibit at the Whitney Museum showing computer-based artwork from maybe the 90s, and the amount of work they had to do to resurrect software systems from computers not even 30 years old was mind-boggling.

It's very hard to insulate yourself from the fast creep of technology where things quickly go from cutting-edge to obsolete and unusable. Try getting a VHS to play these days. It's a real challenge unless you have the right tools.

Maybe we can hedge against this by hoping web browsers exist for a while. But looking at the legal climate today around Google and the Department of Justice, with Chromium powering eight out of the ten most popular browsers, you start to wonder how long we can sustain this support for the modern web as we know it. This makes me wonder if writing everything in JavaScript and WebGL is the best idea for longevity.

The only real hedge is hoping your art becomes popular enough that people pay to store, archive, and keep it running. Fingers crossed, that's all I can hope for.

I'm simultaneously torn by wanting to hone my craft and dive deeper into technologies I'm building real understanding for, while also being drawn to all the cool things being created, especially in AI. Google's Veo 3 video model just came out, and I've been blown away by what you can do with it.

I feel like I'm halfway always chasing new things, which isn't great for building mastery, but it is fun to explore. It feels like now's the time to engage with these tools more deeply. I've been building conceptual pieces around large language models because the fact that they capture this snapshot of the internet and our collective conscious activity is so interesting to probe and play with.

I did a project with DALL-E 2, which created some of the first convincing text-to-image pictures. Even now, maybe two or three years later, the images look incredibly dated and have a certain style I can't reproduce because the model's been taken offline. It's fun that we're building these little time capsules as we play with these tools because they're evolving so quickly. We're making what we can while we have it and producing these artifacts.

Before I discovered computer-based artwork, I became really interested in street photography, particularly the work of Daniel Arnold, a street photographer based in New York. He takes pictures that show some of the most out-of-this-world bizarre scenarios that are completely unscripted, just found by walking through Manhattan for 12 hours a day.

The first time I saw these pictures, I came in expecting art to be all about beauty and showing things that are captivating to behold. His pictures looked very grungy – weird angles, kind of blurry, flash sometimes very apparent. My initial reaction was that these weren't very interesting.

But the moment I dug deeper and tried to see what he was trying to say, they became infinitely interesting. I was living in New York at the time, and these things were happening on streets around me. They felt very quintessentially New York, but also alien because they depicted a degree of absurdity you'd be hard-pressed to find yourself.

This motivated me to buy a camera and try street photography myself, which really heightened the experience because I realized how difficult it was to produce photographs like that. They don't look very pretty on a basic technical level, but there's incredible challenge in framing, getting close enough, finding these moments, and discovering hilarious juxtapositions.

This really reshaped how I look at art. It's all about creating or telling a narrative that goes deeper than the first half-second of immersion – creating something that pulls you into a world you want to spend more time in. That reframing has stayed with me and been a definite guiding force throughout my computer-based artwork.

I fall strongly into the second camp in terms of intentions, but there are all these little connections I can make between different parts of my career that tie it all together. My approach is about applying effort to what's in front of me and not being too planned or particular about exactly what I do, as long as it's something I put energy, time, and investment into.

That's led to working on various things at Google, doing research in undergrad, and now doing computer art full-time. By not being too particular about crafting a big narrative, I feel like it creates far more interesting possibilities. Reality is almost always stranger than fiction, and the kinds of things that can happen if you allow yourself to be open to it are always more rewarding than being very prescriptive.

Of course, nothing's really disconnected from anything else. The street photography interest has certainly followed me into generative art. I wouldn't have been able to draw that connection at the time, but the fact that I did both certainly drives the work I do today.