Experiment: GenAI and Lyric writing

Overview: Lyric writing has always been tough for me. Inspired by David Bowie’s “cut-up method”—cutting and rearranging snippets of text—I wanted to try this approach digitally using AI. Physically cutting up paper wasn’t practical in my home studio, so I recreated the method digitally, hoping AI could help me generate scratch lyrics and break through creative blocks.

View Project

Setting the Stage: Previous attempts at AI lyric generation with tools like ChatGPT or Gemini felt too rigid and lacked my personal touch— no matter how long I spent tweaking prompts. I wasn’t finding it inspiring or productive. I researched methods by songwriters to see how they went about their creative process to find a new approach that offered a way to overcome that.

David Bowie used a variation of “the cut-up method”  – first explored by the Dadaist in the 1920’s and popularised by William Burroughs in the 1950’s – to generate surprising and unique lyrics, unlocking new perspectives in his writing.

I felt this could be a good fit to modernise with GenAI, aiming to create a continuous stream of lyrics that I could chop, change, and inject my own creativity into. 

The Process: I built a simple web app, called BOW-[AI]-E, to create a digital version of the cut-up method. Key features included:

  1. Generate or Input Text: Generate lyrics using AI or input your own.

  2. Chop into Phrases: Split the text into smaller snippets (3-7 words).

  3. Rearrange on Canvas: Drag and drop snippets to experiment with new lines.

  4. Export: Save your new lyrics as a text file.

I used OpenAI’s API for lyric generation and quickly iterated on the design:

  1. Custom Prompts: Allowing custom prompts for more control over theme or genre.

  2. Content Cleanup: Removing unnecessary formatting like VERSE or CHORUS tags.

  3. Flexible Cut-Up: Options to cut by phrases, sentences, or individual words.

  4. De-Dupe Feature: Hiding repeated phrases to make editing easier.

  5. Edit Mode: Enabling direct editing of snippets for more personal input.

Results: The tool helped me generate scratch lyrics fast, giving me creative control while using AI as a catalyst. I used the AI outputs as inspiration, refining them to suit my songs. This hybrid approach of AI assistance and personal creativity allowed me to overcome writer’s block effectively.

Reflection: My goal was impact over perfection—spending just four hours resulted in a tool I’ve used repeatedly. BOW-[AI]-E felt genuine: not just copying AI output but building on it to create something personal. I also logged ideas for future improvements:
– Aaving canvases.
– Adding version tracking. 
– Clustering phrases into chorus/verse/pre-chorus etc.
– Collaborative multi-editor mode.

Conclusion: BOW-[AI]-E showed me that AI can be a powerful co-creator. By generating raw material and then editing and rearranging it, I kept my creative voice intact while benefiting from AI’s unique suggestions. If you’re stuck with lyric writing, I’ve included a link to the (very) rough and ready tool here – go play!

Privacy Settings
We use cookies to enhance your experience while using our website. If you are using our Services via a browser you can restrict, block or remove cookies through your web browser settings. We also use content and scripts from third parties that may use tracking technologies. You can selectively provide your consent below to allow such third party embeds. For complete information about the cookies we use, data we collect and how we process them, please check our Privacy Policy
Youtube
Consent to display content from - Youtube
Vimeo
Consent to display content from - Vimeo
Google Maps
Consent to display content from - Google