The people building AI expect machines, not people, to do the reading. Here's the argument for what that means — as essays to read, and as a podcast to hear.
The full position paper: why the most valuable documents are the ones nobody reads end to end — and what changes when the archive becomes answerable. Grounded in the public thinking of Karpathy, Altman, and Huang.
Read the paper →The shift from reading your documents to asking them, argued in full — from why storage tools can't become answer tools to what verification really costs.
Read the essays →The same thinking as deep-dive audio conversations — six episodes on what it means when machines, not people, do the reading.
Hear the series →The product behind the thinking reads your whole archive so you can ask it anything.
Start free