Large language models show narrowing views on climate topics over time
Narrowing the Horizon: Quantifying Topic Saliency Shifts in Generative Monoculture
Artificial Intelligence
Summary
Large language models are computer programs that help us find and share information, but they can start to show less variety in the ideas they talk about over time. The authors created a way to check which topics or viewpoints become more or less popular in different versions of these models. They tested this with climate change discussions and found that newer models tend to focus less on diverse solutions. Their work suggests using specialized models can keep a wider range of ideas available.
What this means in practice
- •For ai developers: Track which topics lose or gain attention in different language model versions to maintain balanced outputs.
- •For content moderation teams: Monitor topic shifts in AI outputs to prevent unintentional biases from dominating online discussions.
Authors
Oriane Peter, Elena Simperl, Kate Devlin
Abstract
As Large Language Models (LLMs) become central to how we access and share information, they play an increasingly powerful role in shaping global knowledge. However, as these models evolve, their outputs risk converging into a \textit{generative monoculture}, where the diversity of perspectives they represent narrows over time. Studies at the model level often fail to pinpoint which specific topics or viewpoints are being marginalised or amplified in this process. In this paper, we introduce a method to measure shifts in topic saliency across model families, tracking what gains or loses prominence during post-training. Applying this approach to a case study of climate change discourse, we demonstrate how homogenisation affects the representation of diverse solutions across different models. We also test interventions to counter this trend, showing that specialised models can help preserve a broader range of perspectives. This underscores the importance of monitoring topic saliency to diagnose the risks of monoculture and to ensure AI systems reflect a pluralism of ideas. Data and Code are accessible \href{https://github.com/oriane/topic_saliency_shift}{here}.