---
title: "Why is AI-generated language so hard to understand?"
slug: why-is-ai-generated-language-so-hard-to-understand
url: https://listedarticles.com/articles/why-is-ai-generated-language-so-hard-to-understand
canonical_url: https://arxiv.org/abs/2610.29162
content_type: research
language: en
published_at: 2026-10-21T00:00:00.000Z
updated_at: 2026-10-25T08:12:56.177Z
author: "Thomas Hikaru Clark, Matias B. Martin, Kyle Mahowald"
authored_by: human
publisher: "arXiv"
publisher_url: https://arxiv.org/
topics: ["Large Language Models", "Natural Language Processing", "AI Research"]
license: all-rights-reserved
word_count: 220
reading_minutes: 1
citation: "Thomas Hikaru Clark, Matias B. Martin, Kyle Mahowald, arXiv. \"Why is AI-generated language so hard to understand?.\" 21 Oct 2026. https://arxiv.org/abs/2610.29162 (all-rights-reserved)"
---

# Why is AI-generated language so hard to understand?

> Thomas Hikaru Clark, Matias B. Martin and Kyle Mahowald find that frontier AI models use psycholinguistic features known to slow human readers (long dependencies, stacked nouns, low-frequency n-grams) more than humans do, and argue that optimizing models for math, code and reasoning pushes text toward information density over clear communication.

## Abstract

There is a widespread sense that AI-generated text has recently become harder to understand. In this study, we investigate a set of psycholinguistic features that are known to correlate with human processing difficulty (e.g., object relative clauses, long syntactic dependencies, low-frequency n-grams). Using a novel corpus of AI-generated text across diverse domains, we find that AI models use these difficult features more often than paraphrases generated by round-trip translation. Compared with human writing, all nine frontier and near-frontier models use more inanimate subjects with animate verbs and more low-frequency n-grams, and most also use longer dependencies, more stacked nouns, and more reduced object relatives in conversation; we also demonstrate substantial variation across model families. We propose and test a hypothesis as to why: while human language is optimized for successful communication, AI-generated text is increasingly optimized for other goals like math, code, and reasoning, in which there is a pressure for *information density* rather than successful communication. We provide a proof-of-concept for this hypothesis by fine-tuning a language model under a length penalty, which leads to a divergence from human-like linguistic behavior. Our findings highlight the axes of misalignment between linguistic processing demands in humans and LLMs, and point toward possible remedies.

Subjects: Computation and Language (cs.CL); Artificial Intelligence (cs.AI). Licensed under [CC BY 4.0](https://creativecommons.org/licenses/by/4.0/). Full paper: [arXiv:2610.29162](https://arxiv.org/abs/2610.29162).