That is, who is on such a team, what do they do, how do they do it, what does it coest, and why is it important — which is to day, how do you measure its value?
So here's my persistent nagging question: How do we make sense of the new terrain of AI when the AI is being created within organizations that make it tricky to wander and wonder expansively?
Is it by imagining new possibilities and exploring them in playful, thoughtful, and deeply imaginative ways? And for some reason, poetry seems to be a part of that exploration?
Could that be? And if so — why is that?
I mentioned in the previous post that I once heard a poet say that "the primary purpose of poetry is to keep the vision of childhood alive into adulthood." I think this means that poetry helps us maintain our imaginative connection to the world and to ourselves. So what's that got to do with AI?
Could it be that it is like this enticing, weird, eery new terrain that we don't yet understand, and poetry is a way of exploring it in a way that is not rational, ordered, or logical? And we need some of that non-rational, non-logical, non-ordred mechanisms for sense-making that revert us to a kind of childlike state of, variously, wonder, awe, amazement, confusion, dread, and fear of things that go bump in the night and creep out of your bedroom closet in the middle of the night?
I once overheard a poet say, “The primary purpose of poetry is to keep the vision of childhood alive into adulthood.”
That line stuck with me. Poetry, at its core, preserves the imaginative sense of wonder and possibility we had before we learned to see the world in conventional, adult ways. It keeps the world strange and alive.
So what’s that got to do with AI? Because I really want to know.
You have Matt Webb’s Poem/1 clock project. It’s a clock. Except that rather than simply doing clock things like showing the time, it generates a rhyming poem so you look at it and — behold! — a poem emanated from some LLM.
Then there’s the Poetry Camera. Click what we used to call the shutter button — which used to mechanically articulate a complicated mechanical assemablage that would open and close a curtain-like shutter contraption for a very precise and nearly instantaneous moment to let light hit a chemically-coated piece of film — and instead of all that mishegoos, you get a poem rattled out on a little embedded printer..again delivered to you based on what an LLM thinks it sees in the frame.
A project I did a bit more than a year ago called PoemOS — which was a playful exploration of AI-generated poetry using Python programming. This was part of my AI Designed Fictions Research Studio — I prototyped this idea of a PoemOS Processing Unit — a software contrivance that was largely about exploring the idea of a poetry engine that could generate poems based on various inputs, including images, text, and other data.
It was a playful exploration of how AI could be used to create poetry in new and interesting ways. It was also just strange to see it work..at first. It was like what I imagine it must have been like to see luggage with wheels for the first time. Cool, and then it’s like..“oh, what’s the big deal, anyway?”
Today, over a year and a half later, seeing an LLM issue a poem is not that strange at all.
In fact — it’s almost boring. But this is how we explore the new and unknown. A learning-by-making exercise.
It’s not commercial. It’s provocative, generative, exploratory. Which is precisely the kind of work that I think is necessary to make sense of this new terrain of AI — and the kind of work I would love to see more of within the companies that are driving AI forward. In fact, I would argue that such companies need to have expeditionary teams doing this kind of Cultural R&D work and I’d love to help set up and lead such teams.
That’s all to say that I was also reminded of a project from 21 years ago — in 2004(!) — trying to make playful sense of the new pervasiveness of WiFi, I contrived a project I called WiFiKu for Christina Ray’s “Conflux” psychogeography festival in NYC. It was a program running on an PowerBook that ingested WiFi network names that I then attempted to manually conjure into proper Haiku poems.
These projects are all echos of a kind of exploration of possibility within domains and terrains for which there is no conventional sense of what is possible. So..we go to some form of sense-making that is not rational, ordered, logical, and often seen — wrongly, I would argue — as frivolous or impractical. (Or “you’re just fr**king around” as an unimaginative boss once said to me.)
These and many, many other projects — as previously mentioned, for example, Sascha Pohflepp’s Blind Camera from way back in 2006, for example — are ways of casting our imaginative consciousness into new terrains of possibility — and doing so in a material way that is tangible and experiential.
Because none of these projects/products/explorations make sense in the conventional way that a camera, or clock, or even a poem makes sense. They are playful, thoughtful, and deeply imaginative explorations that are meant to introduce a challenge to convention, and bring new ways of seeing, sensing and being in the world.
They are also kinds of proof points on the value of this kind of deeply engaged, tangible form of ‘Cultural R&D’ and an argument for installing expeditionary teams like this within the companies that are driving AI forward.
Having spent 17+ years creating software and hardware products with this kind of expeditionary consciousness, I believe such teams need someone who understands both the imaginative potential and practical realities of bringing speculative work into organizational contexts.
See Also