The weirding of design: thoughts on #AIRetreat

I’m on an anticonvergence swing. The rush to systematise, to codify, the release of yet another manifesto or code of ethics – it all tires me out. These attempts feel premature at best; at worst they feel like vehicles for status and positioning, not genuine impact.

I was worried #AIRetreat would trace a similarly prescriptive path, and I arrived ready to play the role of candid saboteur. Two and a half days isn’t enough for twenty people to reach consensus on something as complex as artificial intelligence. Instead of a doomed attempt to plot an accurate map, I wanted us to tell stories of our respective hometowns, the origins of our myths. I wanted to hear which hills smoulder with the smoke of dragons.

Many of the attendees were wrestling with professional difficulties and neuroses. Existential quandries hung in the air. I know well that events like this risk looking elitist, self-congratulatory – but believe me, there was vulnerability. Souls were bared.

AI is overwhelming in scale. At times I feel like the triangle player in the orchestra – can I really add anything worthwhile? So we talked about giving ourselves the conviction to contribute but the grace to step back. We threw around our pet ideas to weigh them, to examine their shape, to see how they bounce. 

The old designer hammer-nail combo raised its head at times: Post-Its flew onto the windows, and we couldn't help but categorise. Perhaps more interdisciplinarity would have helped. But of course designers do have something to add to the AI conversation; some human-leaning balance to a field colonised by the technical. And that conversation was fast and deeply intelligent. After some early toe-dipping of theory and labels, we leapt in – sex, ethics, consciousness, existential risks, mundane dystopias. We agreed on the folly of separating technology from its social context – augmented reality, for example, makes this fallacy blindingly clear – and argued whether AI should amplify or alter humanity.

It’s important the public has a say in AI discourse, lest we slip into technocracy. But the cultural tropes of AI don’t help. We need to go past the Terminator/HAL angle, the white plastic humanoid handshake motherfuckers. We need new art, new metaphors, new visual and narrative motifs for AI. Black Mirror does a great job in its millennial-Aesop way, but designers and artists have valuable skills that could further provoke the conversation. Can we, for instance, create compelling visions of not the black mirror but the magic window? Can we sketch out a technology that points outward, exploding the hidden components of our environments and lives, collapsing the distances of capitalism (provenance, labour, energy) and helping people make more informed choices?

In this territory, art movements and speculative prototypes surpass manifestos. I think if we’re to really contribute to the territory of AI, we need weirder design practices. We can’t think of interfaces as deterministic, nor interactions as linear. Designers will have to expand both their inputs and outputs: fiction, posters, plays, and games could play roles as large as products and blueprints. I see more value in the futures toolbox than the usability test.

I have to mention the nature. We met at Juvet, famous from Ex Machina, buried in the mountains of northern Norway. The air was crisp and the changing autumn light threw time out of balance. Kairos ruled and chronos melted away. Yellow and ochre leaves tumbled into the river. We marvelled at the Milky Way and the (faint but undeniable) Borealis. We climbed a great big hill, gulping cold air and pointing at the glacier beyond. Once we descended, we drove just a little further, to Trollstigen. As we rounded the bend, we realised the sheer size of the valley ahead, and it took our breath away.

Cennydd Bowles