Corpus Realism
“It’s easier to imagine the end of human readers than the end of writing.”
Corpus realism is the condition in which it is impossible to imagine any writing that will not eventually be assimilated by machines.
The term was coined by Myles Badger in the August 2026 essay “Corpus Realism”, published in his newsletter Myles Mode. It extends Mark Fisher’s notion of capitalist realism — the condition in which it is easier to imagine the end of the world than the end of capitalism — into the era of AI training data.
“Corpus” serves double duty: it names both the body of work a writer produces and the pile of data that large language models are trained on. Corpus realism describes the collapse of the distance between the two. Like Fisher’s realism, it is a condition rather than a position — something experienced as a horizon of possibility, not a belief anyone chooses. Belief in corpus realism is fully compatible with hating it.
What it is not
Corpus realism is distinct from Dan McQuillan’s AI realism (the feeling that AI’s deployment can never be seriously questioned, only reformed) and from Hannes Bajohr’s post-artificial text (the erosion of the reader’s assumption that an unknown text was written by a human). If AI realism is the inability to imagine a future without AI, corpus realism is the inability to imagine not being assimilated by it — and it survives even in writers who have never used a chatbot, because assimilation doesn’t require participation.
Why “realism”
Following Fisher, naming the condition is a way of pointing out that the necessity it describes is at least partly false. Assimilation-without-consent is a policy default in the guise of physics; the essay argues that what corpus realism forecloses is the collective renegotiation of AI assimilation itself — through consent infrastructure, collective licensing, and data trusts.