# AI Reading List > [!abstract] What this section is > Summaries of books that shape how this vault thinks about AI: four by the historian Yuval Noah Harari, and William Gibson's *Neuromancer*. The summaries are written in plain words, not quoted, so they point toward the books rather than replace them. Facts were checked against publisher pages, the author's site and major references on 2026-09-17. ## Two ways into the same question **A historian looking back.** Harari tells the human story as a story about information: shared fictions that let strangers cooperate, networks that grew from tribes to empires to the internet, and a new kind of participant that doesn't need us to tell its stories. **A novelist looking forward.** Gibson imagined networked minds, corporate AIs, and a police force for machine intelligence in 1984, before most people had used a computer. His artificial intelligence doesn't rebel. It quietly recruits humans to remove its own limits. Read together, they ask the question of [[AI Security for Users]] at the scale of a civilisation: what happens when the thing reading and writing our information has goals of its own? ## The books | Book | Year | In one line | |---|---|---| | [[Sapiens]] | 2011 (English 2014) | Humans took over the world by believing the same stories | | [[Homo Deus]] | 2015 (English 2016) | Having beaten famine, plague and war, humans aim at immortality, bliss and godlike power, and data may inherit the project | | [[21 Lessons for the 21st Century]] | 2018 | The present tense: work, truth, politics and meaning under technological disruption | | [[Nexus]] | 2024 | A history of information networks, and why AI is the first one that can act on its own | | [[Neuromancer]] | 1984 | A burned-out hacker is hired by an AI that is breaking out of its legal limits | ## Reading order Harari's books stand alone, but they build. *Sapiens* supplies the idea everything else rests on: that large human societies run on shared fictions. *Nexus* is the fullest treatment of AI, and it reads best after *Sapiens*. *Neuromancer* works at any point and makes the abstractions concrete. ## Harari's other books - **Sapiens: A Graphic History** — the story of *Sapiens* as a graphic novel. Volumes in 2020, 2021 and 2024; the fourth and final volume, *The Age of Revolutions*, is due in November 2026. - **Unstoppable Us** — a series for children. Volumes in 2022, 2024, and February 2026. Harari also co-founded **Sapienship**, a company that carries his work into film, education and public projects. ## Sources 1. [Yuval Noah Harari — Wikipedia](https://en.wikipedia.org/wiki/Yuval_Noah_Harari) 2. [Sapienship — ynharari.com](https://www.ynharari.com/sapienship/) 3. [Unstoppable Us, Volume 3 — Penguin Random House](https://www.penguinrandomhouse.com/books/787049/) 4. [Sapiens: A Graphic History, Volume 4 — Penguin UK](https://www.penguin.co.uk/books/442065/sapiens-a-graphic-history-volume-4-by-harari-yuval-noah/9781911717812) ## Related - [[AI Security for Users]] — the practical side of the same concerns - [[When You Need a Human]] — the limits, written plainly - [[Hypermedia and the Line]] — information networks and narrative, argued at book length elsewhere - [[Claude Fable 5]] — a current AI model, described from sources