# The Dynamics of Human and AI-Generated Language: How Semantics Fluctuates across Different Timescales

arXiv: https://arxiv.org/abs/2606.11371 · HTML: https://mlsanity.com/paper/2606.11371/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 10 of its 120 references that are in the map.
- First posted on arXiv 9 Jun 2026.

## Builds on (most cited first, 10 of 10)

- [Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks](https://mlsanity.com/paper/1908.10084/) (2019)
- [Testing AI on language comprehension tasks reveals insensitivity to underlying meaning](https://mlsanity.com/paper/2302.12313/) (2023)
- [Detecting Subtle Differences between Human and Model Languages Using Spectrum of Relative Likelihood](https://mlsanity.com/paper/2406.19874/) (2024)
- [Surprise! Uniform Information Density Isn't the Whole Story: Predicting Surprisal Contours in Long-form Discourse](https://mlsanity.com/paper/2410.16062/) (2024)
- [A Linguistic Comparison between Human and ChatGPT-Generated Conversations](https://mlsanity.com/paper/2401.16587/) (2024)
- [The Harmonic Structure of Information Contours](https://mlsanity.com/paper/2506.03902/) (2025)
- [FACE: Evaluating Natural Language Generation with Fourier Analysis of Cross-Entropy](https://mlsanity.com/paper/2305.10307/) (2023)
- [Identifying the Periodicity of Information in Natural Language](https://mlsanity.com/paper/2510.27241/) (2025)
- [Calculating the similarity between words and sentences using a lexical database and corpus statistics](https://mlsanity.com/paper/1802.05667/) (2018)
- [Is Information Density Uniform when Utterances are Grounded on Perception and Discourse?](https://mlsanity.com/paper/2602.14653/) (2026)

## Abstract (arXiv)

Spoken language, whether produced by humans or large language models (LLM), unfolds over time with varying semantic content. However, we still lack simple, interpretable time-series features that capture how generic versus specific content is distributed over time, and that can be used to compare human and AI-generated speech. We introduce a semantic-timescale analysis pipeline that turns word-level transcripts with timestamps into semantic time-series. For each spoken narrative, we compute (i) semantic specificity using WordNet-based word depth and (ii) contextual similarity using SBERT embeddings and quantify their temporal dependence using autocorrelation-window measures (ACW-0 and related metrics). We then compare original speech to multiple shuffled controls that selectively disrupt lexical identity, temporal order, and word duration. Across human-read autobiographical narratives, TTS readings, and LLM-generated texts rendered with TTS, we find that segments with longer ACW-0 in the semantic time-series tend to contain more generic vocabulary, whereas segments with shorter ACW-0 are enriched in more specific words. These associations are strongly attenuated or abolished when word order and timing are randomized, indicating that ACW-based measures capture non-trivial temporal organization of semantic content beyond static lexical distributions. Our results suggest that ACW-based semantic timescales are a useful family of features for analyzing and comparing the temporal structure of human and AI-generated speech.