Bad AI-Written Christmas Carols

December 23, 2017

AI researcher Janelle Shane experimented with a neural network to generate unique Christmas carols by training it on 240 popular holiday tunes. This intriguing incident raises questions about responsible AI, as the results deviate from conventional carol themes.

This case study serves as an excellent example for the importance of safe and secure AI practices and governance. It demonstrates how even seemingly harmless AI experiments can produce unexpected outcomes, underscoring the need for guardrails to ensure trustworthy AI. For those interested in shaping Project Cerebellum's HISPI TAIM (Govern) initiative, JOIN US here.

Matched TAIM controls

Suggested mapping from embedding similarity (not a formal assessment). Browse all TAIM controls

MEASURE 2.6 — similarity 0.675, rank 1. TAIM detail and related incidents →
MEASURE 2.5 — similarity 0.645, rank 2. TAIM detail and related incidents →
GOVERN 1.2 — similarity 0.641, rank 3. TAIM detail and related incidents →

Alleged deployer: janelle-shane
Alleged developer: janelle-shane
Alleged harmed parties: carollers

Source

Data from the AI Incident Database (AIID). Cite this incident: https://incidentdatabase.ai/cite/62

Data source

Incident data is from the AI Incident Database (AIID).

When citing the database as a whole, please use:

McGregor, S. (2021) Preventing Repeated Real World AI Failures by Cataloging Incidents: The AI Incident Database. In Proceedings of the Thirty-Third Annual Conference on Innovative Applications of Artificial Intelligence (IAAI-21). Virtual Conference.

Pre-print on arXiv · Database snapshots & citation guide

We use weekly snapshots of the AIID for stable reference. For the official suggested citation of a specific incident, use the “Cite this incident” link on each incident page.