Hey everybody! I realised I've read more LLM-gene...
# share-your-work
a
Hey everybody! I realised I've read more LLM-generated text than human-written text over the past year. So, I created a text editor called Grain that renders the process of writing into the type itself (tempo as weight, erasure as residue, sessions as fading), to remind myself of the human on the other side. I also wrote an essay (The Grain of Thought) outlining how I came to this realisation, some French literary criticism on the body in communication, and the experiments that led to the final project. I would love to hear where I potentially went wrong / any feedback :))
❤️ 13
🤍 1
💙 1
d
You might find Expressive Keyboard (

https://youtu.be/0HBw47b2gno?t=20

) interesting, which uses variation in mobile keyboard gestures to induce different text styles. I also really like how the paper is motivated:
[Standardization of text via digital computers has resulted in] a reduction in personalization that had been present in earlier writing systems. Letters are recorded perfectly with stylistic information stored separately, and applied to large, heavily quantized blocks of text. For example, the subtle emphasis encoded implicitly in a continuously varying pen stroke is now simply rendered as a standard italic typeface. Along with a severe reduction in the granularity of control, this approach also discards potentially valuable channels for implicit communication of personal style, mood or emotional state, and temporal or situational contexts. Users can of course edit the font, typeface size and color of the rendered text, but this is necessarily separate from the actual text input.
[...]
Our goal is to increase information transfer in textual communication with an instrument that enables users to express themselves through personal style and through intentional control.
❤️ 2
a
Ooo I love this - thanks for sharing!!
👍 1
f
Beautiful idea! Good use of variable fonts.
❤️ 1
a
Thank you!! :))
j
Ooh, thanks for sharing this work. Great to see more Barthes in tools. @Max Kreminski mentions that DraftMarks explores similar affordances. https://arxiv.org/abs/2509.23505
👀 2
❤️ 1
r
I love this! I make writing tools (eg enso.sonnet.io) and I would love to see this applied to poetry. I've tried encoding the dynamics of typing/writing in my experiments. It reminds me of calligraphy which we subconsciously read through motion (a fast confident stroke will always look different from almost the same shape but drawn slowly) One piece of feedback: decide whether you want to keep formatting (eg bold) on paste.
👀 1
❤️ 3
a
@Jasmine Otto Ah this is so cool - I'll enjoy reading this as weekend reading! While I distance myself from provenance technology in the essay, I think there is a real need for non-judgmental shared symbols / syntax to identify how and where LLMs have been involved in work, communication, and art. Glad to see there are so many people thinking about this. Also, yes to more Barthes :)
@Rafał Pastuszak This is beautiful.. Love the idea of "kind software" btw
❤️ 1
i
Would be super interesting to run a study and see how much you need to add before it is possible to begin recognizing/differentiating the writers solely by how their writing process is reflected visually (it might already!) I'm also reminded of this and this
❤️ 1
👀 1