Skip to content

UX Design

Press Start: What a Platformer Can Teach Us About Design and AI

Martin Dabrowski

Martin Dabrowski

Martin Dabrowski is a Swedish designer based in Stavanger who has taken digital products from idea to release since 2001. He has worked with AI for five years and uses it throughout the design process, from research and ideation to prototypes and finished products he builds himself. Games keep coming back in his work: his career began with a subscription service for competitive Counter-Strike in 2001, and he has previously talked at Hello Stavanger about what products can learn from games.

His view on AI is simple: it is an excellent reviewer and a very poor decider. The direction and the choices remain human.

1 / 2

This talk is a game. But it isn't about games.

I'm a designer and developer, and I built a 16-bit platformer from scratch with AI – without writing a single line of code myself. I play it live on stage, starting on the Stavanger waterfront. Every time I headbutt a question block and grab the egg that pops out, the next part of the talk opens.

The game is just the example. The same approach works for any project a designer is part of, whether it's a service, an app or an internal tool. Along the way, I'll show how AI can support the whole process:

• Research: Is this even possible, and what's a realistic scope? • Ideation: from a vague idea and a quick sketch to a clear concept • Context and references: giving AI something concrete to work with • From design to something that works: building what you design • Critique, iteration and testing: why the first result is rarely the right one, and what still needs a human eye

The point: AI lets designers take an idea much further on their own than before. But direction, understanding and judgment still decide whether the result is any good.

Unlimited lives. One shot on stage.