The hand-wringing around A.I. and the state of the writing profession is a familiar refrain, but it’s worth considering the potential of A.I. as a creative collaborator rather than a mere tool.
In her The New Yorker article, Anna Wiener explores this idea, suggesting that A.I. can be more than just a gimmick in writing. She references early LLM-generated texts like Vauhini Vara’s “Ghosts” and Stephen Marche’s “Death of an Author,” which demonstrate the potential for A.I. to assist in the creative process rather than replace it.
Wondering about these possibilities and what a future of the LLM or A.I. collaboration might look like is the motivation behind the Vibewriter and Ghostwriter projects.
While I have tried writing (of a sort..I’m not sure I’d call it that precisely) with LLMs early on (cf Android’s Dream of Electric Sheep), I have found that the most interesting and useful applications of A.I. in writing are those that are more like ones where I feel like I’m jamming with A.I. rather than letting it simply generate text. (I’ve used the metaphor of “trading fours” in the past somewhere here.)
The Vibewriter project, for example, comes from a possible future in which there is a felt desire to retrieve writing from the perceived drudgery of just getting words on a page, or synthesizing ideas into text. Instead, this is seen as a creative act. Maybe a bit like a return (by some) to wet photography, or vinyl records, or analog synthesizers, or even typewriters (the mechanical or electro-mechanical kind).
As a design fictional conceit, we can imagine such a future where A.I. took over writing, because we often race to extreme ends of the spectrum. But then there is a desire to reacquaint with the craft of writing and relearn the nuances of language, style, and voice.
In this design fictional future, I imagine that the return to the craft of writing is a bit of a movement. A response that is a bit Punk, and confusing to some, but a way to distinguish the fact that this was written by the hand of a human, and in this way is a more precious artifact.
If I pushed this design fictional conceit further, I can imagine that there has been a diminishment in the capability of the human brain to write, as if that corner of the brain has atrophied from disuse. In this speculation, you might see articles describing this bio-evolutionary retreat; you might see arguments on all sides of the issue claiming that we’ve actually advanced evolutionarily and cognitively as we never needed to write to begin with, and that that part of the brain should go the way of the appendix.
On the other side, you’d have the children and grandchildren of the literatti who are reclaiming the craft of writing, holding seminars and workshops and retreats to remind their adherents to the value of writing and the nuances of its craft, speaking of it as if even the word “writing” has been somewhat lost to the generation who is now rediscovering it (again, in the same way that there has been a return to the analog, to film photography, etcetera.)
These are all speculations, of course.
The impetus for these experiments in software – artifacts I created to help me think through these ideas; artifacts I created because I had a feeling I had to get into conversation with in a material way; because I can make things that help me think and wander into possible new territories and terrains — and we are at the vanguard of a new territory (of writing and creativity, at the very least).
And as regards this new terrain, it is quite expansive. And we are, as is our wont, bringing our old language and ideas about the value and utility of A.I. right into this, and I feel that those ideas are not quite right. They are not quite right because they are too narrow, too limited, and too focused on the idea of A.I. as a tool that wholesale replaces human effort.
Sketchnoting revolutionizes note-taking by transforming complex...
A...
