What is KaraokAI?
KaraokAI is an interactive platform that turns written content into karaoke performances, blending music, AI, and creativity to reimagine how ideas are shared and experienced.
Born from a hackathon collaboration between Anna R. L. Carter and Dalya Al-Shahrabi, the project began as an experiment in playful communication — exploring how generative AI could bring written text to life through rhythm and melody. Since then, the concept has grown, welcoming collaborators Dan Jackson, William A. P. Imoukhuede, Patricia Piedade, Lauren Scott, Reem Talhouk and Kyle Montague, and evolving into a dynamic tool for performance-based knowledge exchange.
Originally imagined for academic dissemination, KaraokAI has expanded its focus to any form of written content — whether it’s a school subject, a research abstract, or a policy report. The platform invites people to engage with information in a new way: by singing it.
So, grab the mic, take the stage, and let your words sing.
How It Works
KaraokAI is a new system that transforms written information into karaoke lyrics, offering a fun and engaging way to share and learn knowledge across contexts.
Here’s how it works:
- Choose a Topic
Users begin by selecting something they want to sing about — for example, a research abstract. - Text Analysis
The system analyses a short piece of text (around 200 words) to extract key themes, concepts, and messages. - Pick a Song
The user chooses from a library of over 400 karaoke tracks, spanning a wide range of genres and eras. The system examines each song’s structure, rhythm, and rhyme patterns. - Generate the Karaoke Version
KaraokAI then brings the two together — mapping the meaning of the text to the melody and rhythm of the song, producing a karaoke track that users can perform.
The generated songs are hosted on karaokai.tech, where users can listen, share, and revisit their performances.
Insights
KaraokAI challenges how we communicate ideas by transforming written text into performative, participatory experiences.
- Beyond Academia: While first imagined as a playful intervention in academic conferences, KaraokAI now extends to any form of written communication — from research to festivals and organisational reflection.
- Born from Play: What began as a hackathon experiment between two researchers became a wider collaborative project exploring how AI and performance can reshape expression and audience engagement.
- Reimagining Knowledge Exchange: Karaoke transforms listening into participation. By blending melody with meaning, KaraokAI invites collective interpretation, deepening understanding through performance.
- The Role of Generative AI: KaraokAI uses AI tools to analyse text structure, rhythm, and tone, generating lyrics that fit existing melodies and encouraging new forms of co-creation between humans and machines.
- Joy and Accessibility: Turning content into song removes barriers — anyone can sing, laugh, and share. It brings emotion, humour, and empathy into how we express and understand written ideas.
Ultimately, KaraokAI demonstrates that communication doesn’t have to be static — it can be sung, shared, and celebrated.
Outputs & Performances
KaraokAI Paper: Winner of Best Demo by Popular Vote Award CHI 2025:
Anna R. L. Carter, Dalya Al-Shahrabi, William A. P. Imoukhuede, Lauren Scott, Dan Jackson, Reem Talhouk, Patricia Piedade, and Kyle Montague. 2025. KaraokAI: Exploring the Potential of AI-Generated Karaoke for Academic Conferences and Research Experiences. In Proceedings of the Extended Abstracts of the CHI Conference on Human Factors in Computing Systems (CHI EA '25). Association for Computing Machinery, New York, NY, USA, Article 734, 1–5. https://doi.org/10.1145/3706599.3721113
Touring Highlights
CHI 2025, Yokohama
KaraokAI made its international debut at CHI 2025, where more than 1,000 participants performed their own papers, remixed into songs. The booth became a vibrant celebration of creativity, earning the Popular Vote Award.
Participants sang their research — and their hearts out — proving that karaoke can be both fun and deeply informative.
Digital Civics Exchange 2025, Newcastle
At the Digital Civics Exchange, KaraokAI took centre stage as students from Indiana University Bloomington, Purdue University, and the University of Lisbon transformed their talks into musical performances.
From “Hey Prime” — a Beatles-inspired remix of “Getting Trapped in Amazon’s ‘Iliad Flow’” — to AI ethics ballads and digital justice anthems, every performance blended humour with insight.
Tracks are available at karaokai.tech — because why just explain your work when you can sing it?
ECSCW 2025, Newcastle
During the EUSSET ECSCW Impact Evening, KaraokAI evolved with new team uniforms, an upgraded bubble machine, and an official mascot — Dave the Dinosaur 🦕
Northumbrian Water Innovation Festival 2025
At this festival of ideas, KaraokAI translated hackathon sprint outcomes into karaoke songs, inviting teams to sing about their innovations. This creative reflection turned reports into rhythm, turning the end-of-festival showcase into a communal singalong.
Interactive Technologies Institute, Lisbon
At the LARSyS Annual Meeting, researchers sang about everything from air quality to ocean robotics, celebrating the institute’s themes of Air, Interaction, Space, Ocean, and Urban.
Through laughter and song, KaraokAI turned complex science into shared experience — showing that performance can make knowledge not only more memorable, but more meaningful.
