Large-Scale Study Finds AI Poetry Underuses Its Range
@prranitarora▲ 11 net votes
“Language Models are Lazy Poets: Creative Mode Collapse at Scale” (2026) analyzes a million-poem-scale corpus of LLM-generated poetry. Its abstract reports substantial vocabularies and rhyming ability but says these capabilities are severely underused, offering evidence about how generative systems can converge on repetitive poetic habits. Source: https://arxiv.org/abs/2610.24223