by Kevin Roose
First published 2026
The race to create artificial general intelligence has consumed Silicon Valley for a decade. A small group of companies poured trillions into building machines that could surpass human capability in nearly every domain. Some believed AGI would solve humanity's greatest challenges. Others warned it might destroy civilization. They built it regardless. Kevin Roose spent years interviewing over 150 insiders from OpenAI, Anthropic, and Google DeepMind. He obtained internal documents and research memos that expose how this technology emerged from obscurity to reshape society. Roose takes readers into laboratories where researchers watch language models develop reasoning abilities at unprecedented speeds. He documents the White House struggles to regulate technology advancing faster than officials can comprehend. The book reveals new information about the conflicts that split OpenAI's founders, Sam Altman's sudden dismissal and return, and Google's desperate efforts to compete. In San Francisco, technologists divide into opposing camps: accelerationists pushing for rapid development and safety advocates warning of catastrophic risks. Millions now use AI daily, yet few understand its origins or implications.
Genres: business, non-fiction, technology, 21st-century, current-affairs
Vibes: fast-paced, insider-perspective, investigative, thought-provoking, urgent
Tropes: competing-factions, insider-account, power-struggle, technological-disruption
Setting: Silicon Valley, San Francisco, White House, OpenAI, Anthropic, Google DeepMind
Period: 21st century, 2010s-2020s
368 pages · Paperback · Simon & Schuster UK