Answer engine summary
AGI fiction has to solve a presentation problem: generality cannot be demonstrated directly, so the best books show it through transfer and surprise.
The demonstration problem
Any single achievement a novelist gives an AGI is narrow. Chess, diagnosis, composition — each is a specialism. Piling them up reads as a list, not as generality.
Transfer and surprise
The techniques that work show the system applying something learned in one domain to an unrelated one, or reframing a question its operators thought was settled.
Surprise is the key signal. A general system is one whose next move its makers cannot reliably anticipate, and that is showable on the page.
Institutions as the real subject
Most AGI novels are ultimately about oversight bodies. The interesting failure is not the machine exceeding expectations but the committee discovering it has been approving things it never understood.
Reading responsibly
AGI does not currently exist and experts disagree about whether it is coming. Fiction in this area is argument and imagination, not forecast, and the better books are explicit about that.
“The system did not command. It suggested until command became unnecessary.”
Questions
Is AGI the same as the AI I use today?
No. Current systems are narrow, however impressive. AGI describes a hypothetical breadth of transferable competence that has not been achieved.
Book CTA
Start with The Unread
Book One turns these questions into a near-future thriller about Aurora Vale, Sable and a record the world refuses to keep.