SYNTHESIS NOTE
Topics›this note

Does AI writing lack the internal appeal to attention that humans use?

Explores whether AI-generated text is structurally missing the constitutive property of human communication — an internal gesture that reaches for and holds the reader's attention, not just inheriting visibility from platforms.

Synthesis note · 2026-04-14

Language used to communicate carries an internal property — an appeal to the attention of the addressee. The writer is doing something to the reader: inviting them in, holding them, asking them to follow. This appeal is not an extra layer added by good writers; it is a constitutive property of writing-as-communication. Speech act theory and Goffman's interaction order both name this from different angles. Without the appeal, the utterance is not addressed to anyone in particular even if it has surface form.

AI-generated posts inherit attention from the platform context — they are visible because the algorithm placed them in front of users — but they do not perform the appeal internally. The text does not reach for the reader. It states. The platform supplies the attention; the writing does not seek it. This is what produces the aloofness readers report about AI text: not a stylistic flatness that could be stylistically corrected, but a structural absence of the gesture that constitutes communicative writing.

The mechanism follows from how the model is conditioned. Generation maximizes likelihood of next tokens given the prompt context. There is no audience model beyond the prompt-as-stand-in-for-audience. The model can produce attention-seeking surface forms (hooks, questions, dramatic openings) when prompted, but the generated text is performing the surface of those forms without the internal appeal that the surface forms are markers of in human writing.

This is distinct from but compatible with Why do LLMs produce such different writing in chat versus posts?. The chat register has a kind of pseudo-appeal because it is conditioned on conversational patterns that include attention-seeking moves directed at the prompter. The post register lacks even that — the prompter is the only addressee, and the prompter is not the audience. So the post register reaches no one in particular, which is precisely the absence the source describes as aloofness.

Inquiring lines that read this note 98

This note is a source for these research framings, grouped by the broader line of inquiry each explores. Scan the bold lines of inquiry; follow any specific question forward.

Are AI-generated articles systematically disadvantaged in search ranking and user engagement? Can readers reliably distinguish AI-written text from human writing? What enables conversational agents to guide rather than just respond? How do users confuse explanation quality with actual system accuracy? Can artificial systems establish authority in domains requiring expert judgment? Can AI systems participate in genuine communication or only simulate it? How does AI-generated content create social proof without authentic interaction? What determines AI's persuasive power and how can it be detected or mitigated? What are the fundamental limits of prompting for language models? How reliably can humans and AI detectors identify machine-generated text? How do writers navigate authorship and delegation with AI? How do interpretive frames override surface features in text comprehension? Can AI systems achieve real improvement without external human feedback? How do philosophical assumptions about AI consciousness affect practical harms and design? What distinguishes genuine communicative competence from surface language performance? Does disclosing AI authorship change how audiences evaluate the writing? How should AI agents balance proactive engagement with conversational respect? Can LLMs distinguish between linguistic form and semantic meaning?

Related concepts in this collection 3

This note in its neighbourhood — explore the map, then jump to a related concept in the list below.

Concept map
19 direct connections · 155 in 2-hop network ·medium cluster Open in graph ↗

Click a node to walk · click center to open · click Open in graph to see this note in the full knowledge graph

your link semantically near linked from elsewhere

Related papers in this collection 8

Papers most semantically related to this note, ranked by cosine similarity in the embedding space.

Original note title

human social media writing makes an internal appeal to audience attention that AI writing structurally lacks