Does post length explain why AI-flagged text jumped on Medium but barely moved on Reddit, or something else?
Does length of post explain differences in AI rates across formats?
This explores whether the gap in AI-generated content between platforms (for example, long-form Medium essays vs. short Reddit comments) comes down to how long the posts are, or whether something else drives it.
This explores whether post length explains why some platforms show far more AI-written content than others. The short answer: the collection points toward length but never tests it directly, so the honest position is that it's a plausible explanation, not an established one. The clearest data comes from a 2.4-million-post study. Detector-flagged AI content on Medium and Quora jumped from about 2% to about 38% between early 2022 and late 2024, while Reddit barely moved, going from 1.31% to 2.45% Is AI-generated content rising faster on some platforms?. Medium and Quora reward polished, essay-length answers, and Reddit runs on quick conversational replies. The split fits a length story, but platform, length, audience and incentives all change together, so the data can't tell you which one matters.
The economics are a useful side angle. A study of a Chinese short-video platform found that AI creators keep up with human creators by posting more, even though viewers like their content less Can AI creators match human creators through posting volume alone?. Length may matter less as a feature of the text than as a measure of effort. The more work a format takes, the more an AI draft saves. A 1,500-word Medium post is costly to write by hand, while a two-line Reddit reply costs almost nothing, so there is little reason to hand it to a model. On this reading, 'length' is really effort saved.
What happens after drafting may also play a part. Writers edited AI-generated paragraphs only 23% of the time, and their edits left the text about 96% the same Do writers actually edit AI-generated text before publishing?. Long-form posts drafted with AI are therefore likely to be published nearly untouched, so they keep the style a detector picks up. The style also shifts in predictable ways: AI help pushed writers toward sounding more confident, more extreme and more polished on every one of 29 measured traits Does AI writing assistance change how readers perceive the writer?. Those signals have more room to pile up across a long essay than in a quick comment.
The corpus is missing several things. No study holds the platform fixed and varies post length. None checks whether AI detectors behave differently on short and long text, which is an obvious alternative explanation for the low Reddit number. Readers also seem to care less than you might expect: AI-flagged Medium posts got roughly half the likes of human ones (69 vs. 128), which the authors call a fairly small gap Do readers engage less with AI-generated social media posts?. If you want to dig further, the question to ask is less 'is it length?' and more 'is it the effort the format demands, and can the detector even see AI in short text?'
Sources 5 notes
Analysis of 2.4M posts using the OSM-Det classifier found AI attribution rates jumping from ~2% to ~38% on Medium and Quora between January 2022 and October 2024, but rising only from 1.31% to 2.45% on Reddit. The surge began in December 2022.
AIGC creators on a Chinese short-video platform uploaded more videos and achieved comparable total engagement to human creators, even though consumers showed lower valid-view and full-view rates for AI-generated videos. Lower marginal effort in AI production enables this scale-over-preference dynamic.
Writers edited AI-generated paragraphs only 23% of the time, with edits averaging 96% similarity to the original. This means AI's opinionated and distorted voice propagates with minimal human filtering before publication.
A study of 2,939 writers and 11,091 readers found AI assistance shifted every tested dimension—29 total—toward extremism, confidence, quality, agreeableness, and perceived privilege. Distortions were statistically significant and directional, not random noise.
AI-labeled posts on Medium averaged 69.15 likes versus 127.59 for human-labeled posts, with similar gaps in comments across all follower groups. The paper calls this gap relatively small and suggests AI content still appeals to users.
Papers this line draws on 8
The research behind the notes this line reads — ranked by how closely each paper relates.
- Understanding Reader Perception Shifts upon Disclosure of AI Authorship
- Are We in the AI-Generated Text World Already? Quantifying and Monitoring AIGT on Social Media
- Is it Cake or is it AI? A Systematic Review of Human Uncertainty in Distinguishing Generative Artificial Intelligence Content
- AI Now Writes as Many Online Articles as Humans
- Measuring and Mitigating Persona Distortions from AI Writing Assistance
- Penalizing Transparency? How AI Disclosure and Author Demographics Shape Human and AI Judgments About Writing
- What Influences Readers' and Writers' Perceived Necessity of AI Disclosure?
- "It was 80% me, 20% AI": Seeking Authenticity in Co-Writing with Large Language Models