Pre-literate refers to humans without written language — cultures for whom the technology of writing doesn't yet exist. Illiterate means unable to read in a world that expects you to. A post-literate world is one where nearly everyone can read, but fewer and fewer people do it deeply, the way we had for the last three hundred years: by reading books.
Deep reading has always been a practice of an elite minority. You not only have to have the time, but you also have to have the access and the desire. The percentage of Americans who read for pleasure on any given day decreased from 28% in 2004 to 15% in 2023. Orality, on the other hand, is always at 100%. We are oral creatures by nature, regardless of whether we are pre-literate, illiterate, or literate.
The recent popularization of AI was made possible because it tapped into this orality, by exploiting the interface of conversation. But this new form of orality that is emerging is different. It's bathed in literacy, and it’s generated by numbers. Let's call it synthetic orality. It feels familiar, but something has fundamentally changed. The thing we are talking to, is it a machine? A being with consciousness? A stochastic parrot? Or are we talking to ourselves?
And what happens to knowledge itself, if we are no longer reading books, when we can outsource the critical work of comprehending difficult texts, when they can be reconstituted for us by AI into bullet point summaries? What do we lose, and can we reclaim this activity (deep reading) in some other form? Or is it necessarily tethered to the form of the book, the way memory is tied to ritual?
This blog is where I think through these questions. I'm also working on a book about it — called After the Book — to be published inside Virgil, an iOS reading app I'm building with the help of Claude Code. (That’s right. I’m using the thing to examine itself.)
I'm Irwin Chen. Follow along on Mastodon.