Language Understanding and Pragmatics Psychology and Social Cognition

Does AI-generated content mirror oral culture's knowledge patterns?

Walter Ong's framework for oral versus literate cultures may describe how AI content functions on social media. Understanding this parallel could explain why AI discourse feels fundamentally different from print-era knowledge.

Note · 2026-04-14
What do language models actually know? What happens to social order when AI removes ritual constraints?

Ong's Orality and Literacy (1982) drew sharp contrasts between knowledge as it lived in oral cultures and knowledge as it lived in print cultures. Oral knowledge was performative (existed in the act of speaking), additive and aggregative (piled on rather than analytically subordinated), close to the lifeworld (situational rather than abstract), participatory (not objectively distanced), and homeostatic (the society sloughed off memories to live in a self-equilibrating present). Print knowledge was the inverse: objectified, analytically structured, abstract, distanced, archival.

AI-generated content on social media exhibits the oral pattern with eerie precision. It is performative — exists in the act of generation, not as a durable archived object. It is additive and aggregative — AI piles connections rather than building rigorously subordinated arguments. It is close to the user's lifeworld — contextual, prompt-driven, situational rather than categorical. It is participatory — the conversational interface frames it as response rather than as treatise. It is homeostatic — outputs are disposable, regenerable, not laid down as permanent record.

Ong already had a category for this: secondary orality, which he used for radio and television. AI is secondary orality pushed further — orality without a speaker, conversation without a participant, performance without a performer. The oral pattern of knowledge persists; the embodiment that gave the oral pattern its anchor is gone. Where is the speaker when AI produces speech? is the next-level claim.

The match is not coincidence. The same medium-properties that produced the oral pattern in archaic cultures — knowledge as flow, value as performance, low cost of regeneration — are reproduced in AI by the architectural facts of generative production. Tokens are generated, not stored. Outputs are performed, not archived. The cultural form follows the technical form.

This has consequences for how social media discourse will evolve under AI saturation. Intuitions calibrated to print-era discourse (the article, the argument, the position-paper) will misfire because AI is not producing print-era artifacts even when the surface form looks print-like. The relevant comparison set is oral cultures, with their distinct strengths (immediacy, participation, situational fit) and distinct failure modes (loss of analytical rigor, homeostatic forgetting, dependence on speaker authority).


Source: Tokenization of Intelligence - Theoretical Extensions

Related concepts in this collection

Concept map
12 direct connections · 92 in 2-hop network ·medium cluster

Click a node to walk · click center to open · click Open full network for a force-directed map

your link semantically near linked from elsewhere
Original note title

AI on social media is a return to orality — performative additive situational and disposable like Ong's oral culture