The Future History of AI

Decades before the first large language model, science fiction was already running the experiment. Writers imagined thinking machines in detail — how they'd reason, where they'd fail, who they'd serve, and what they'd cost us — and a surprising amount of it is arriving on schedule.

This blog reads those classics closely. Each post takes a famous work — 2001, I, Robot, Neuromancer, Ex Machina, and the rest of the canon — and asks two questions: What lesson was it really encoding? And what does it tell us about the AI we're actually building right now?

The best science fiction was never about predicting gadgets. It was a simulator for consequences — a way to test-drive a future before we had to live in it. We're living in it now. These stories are the field notes we wrote in advance.

Start with the analyses, or read more about this project.