A Cyborg Digital Writing Primer
- Cyborg writing refers to how our own writing is increasingly shaped by digital writing technologies such as spell and grammar checkers, predictive keyboards and suggested texts, generative AI, and bespoke engines for text generation. Cyborg writing can result in writing designed for the page (printed or virtual) or in digital writing.
- Digital writing refers to writing that embraces the potential digital media offers to produce generative, interactive, multimodal, kinetic, hypertextual, textual works that engage digital culture to go beyond the limitations of the page (printed or virtual).
- Electronic literature is literary digital writing. As an older term in the field, it includes practices that aren’t strictly digital writing, such as sound and video poetry in which language is not presented in written form.
Tutorial #1: How to Hack a Digital Work
Directions: The first step is to select any standalone work (that is, not using external libraries) using Hypertext Markup Language (HTML), or one of the recommended ones below:- Gram’s Fairy Tales by Nick Montfort
- The Two by Nick Montfort
- Taroko Gorge by Nick Montfort
- A House of Dust by Alison Knowles and James Tenney
- Leonardo Flores’ Generative Template by Leonardo Flores
- Any of the works published in Taper.
All of these works are published with open licenses that allow use and modification, with attribution.
Process
- Open the work, linked to above in your browser, and save it as an HTML file in your computer (smartphones and tablets are not recommended for this).
- Open the saved file with a code or text editor (I recommend Visual Studio Code).
- Open the same saved file in a browser. Whenever you make changes to the original file, save it in the editor, and reload it in the browser to see the changes.
- Modify the code as you see fit, being careful not to break it by missing a comma, quotation mark, semicolon, etc. Take small steps in making changes.
- When you’re satisfied with the text the work is generating, give it a title, modify the aspects of the source code that display your title and name, modify the style sheet as you see fit, and rename the file as something that is yours. I recommend writing an artist’s statement in the source code documentation in which you credit the original work and any AI systems you used to assist in coding.
- See if the code editor identifies the problem by making the line number red.
- Open the developer tools in your browser and look for the error messages, which will tell you where the problems are by line and space number.
- Copy-paste your code into an AI system and ask it to fix it. Describe the problem in your prompt or simply explain that there’s a problem with the code and ask to fix it without modifying anything else.
- Use undo (Control/Command + z) until you get to an earlier version of your document that works.
Tutorial #2: How to Use AI to Create a Digital Work
AI can significantly reduce the access barrier to creating new frameworks for digital writing, but you do need to know some programming basics and a general familiarity with the conventions of the programming language you will use and debugging tools. For the purposes of this exercise, I will focus on HTML (including CSS and JavaScript).Preparation
Before you start prompting, think of what framework for digital writing you would like to create. Here’s an example I will illustrate after describing the process:A word (or phrase or poetic line) appears in the middle of the screen. When the user clicks on it, it changes to a different word. This allows you to write a sequence of words that can complete a sentence, tell a story, or deliver a poem.
Process
- Create a new document with a code editor (I recommend Visual Studio Code) and give it a simple name in lower case, followed by the suffix .html (such as sequence.html).
- Open Claude, Gemini, GitHub Copilot, Chat GPT, or AI system of choice. If you prefer not to use large corporate frontier models, you can use an open, downloadable model that runs in your computer using software like Ollama or AnythingLLM and then download a model that can run in your hardware configuration. I recommend Qwen and Gemma for coding.
- Prompt the AI to generate HTML code to produce the results you want, starting with a simple yet descriptive prompt.
- Copy-paste the generated code into your document.
- Save the document and then open (or reload) in your browser to see the results.
- If you don’t like the results or would like to modify them, repeat steps 3–5.
- Once you’re happy with the “engine,” modify the code directly in your editor so that it says what you want it to say. Repeat step 5 after each modification.
Note: You can create a smoother workflow by integrating any LLM into VS Code (GitHub Copilot offers native integration, the others require a plugin or extension), which allows you to create agentic workflows and have suggested code autocompletion as you work directly on the code.
If the code breaks you have some options for debugging:- Describe the problem to the AI system and ask it to fix it.
- You might need to provide context for the AI system on the error message identified by the code editor or browser debugging tools.
- Use undo (Control/Command + z) until you get to an earlier version of your document that works.
Example
Prompt: “Generate HTML code to produce a page in which there’s a word centered on the screen. When the user clicks or touches the word, it changes to the next word in a sequence of 10 words. Use lorem ipsum text for the words.” See what the output looks like—different systems will produce slightly different results—and request changes so the output meets your vision for the piece. As you do this, take a moment to read the code and begin to learn how it operates. There are many changes you can make manually, but be sure the AI system knows the changes you’ve made by copy-pasting the updated code, attaching the updated saved file, or through code editor integration. Some final touches:- Change the title for the piece by modifying the line
<title>Word Sequence</title>to whatever title you want to give it. - Add some documentation to give credit where credit is due by inserting the following code in a new line after the title tag and modifying it with your details. (Note that I have included a short version of the MIT license):
<!-- "[title]" Copyright (c) [year] [author] Copying and distribution of this file, with or without modification, are permitted in any medium without royalty provided the copyright notice and this notice are preserved. This file is offered as-is, without any warranty. --> - Now, play with this simple code and develop it in a variety of ways, such as:
- Randomize the sequence of words.
- Have the click affect the font size, so the words are larger, smaller, or randomly sized.
- Make the word’s color or background color (or both) change with every click.
- Make the words appear in a random location on the screen with every click.
- Make the user be able to drag the words.
- Make the words move in a random direction and fly off screen, or bounce when it hits the edges.
- While you’re at it, make the words accumulate so the whole phrase is on screen.
- Make the words grow or shrink in size over time and or fade.
- Do whatever else you can think of.
- A word of advice: add one feature at a time and work on it iteratively until it functions as desired before adding another feature.
(Updated on July 3, 2026)
Categories: Resources