I intentionally avoid writing another book about the technology as seen from Silicon Valley or the City of London, or elsewhere. This is the third book in a series. In the first I went back to the oldest stories people tell, about made minds and forbidden knowledge and asked what they already knew about our fear of thinking machines. In the second I argued with a famous book about power, and set against its rules a set of practices for a time when much of what once took cunning can be done by software. Both books were about how to think. This one is about what we should observe.
Most of the research on AI use is done in a handful of wealthy countries. That, is the reason I include an "Around the World" section in every chapter herein. It is where I look for evidence from outside the United States and Europe, where I could find it: Brazil's national survey of businesses, Japan's communications ministry, the Masakhane community that builds language tools for African languages, India's national translation mission, the Philippines' outsourcing industry, Kenya's courts, surveys that reach Bangladesh and Sri Lanka and Nigeria. In some chapters the section is strong. In others it is thin, and I say so in the text. Chapter 8 has no primary case from the Global South. Chapter 10 has no African, Arab or Latin American barometer item on AI. Chapter 6 has a press report on the Philippines but not the industry's own figures. I have not hidden those gaps behind a general sentence about "emerging markets." A reader in Lagos or Manila or Dhaka deserves to know exactly where the evidence stops.
A chatbot gave a grieving man the wrong advice about his airfare. A welfare agency sent out more than 20,000 automated debt notices a week. Most American teenagers have tried an AI companion. None of it came with an announcement. It just showed up. AI: 10 Shifts follows ten changes that reach ordinary life, from software that answers to software that acts, from a tool to a companion. Each chapter opens in a real place, weighs the best evidence from around the world, and sets it beside an old story that faced the same question. No forecasts, no hype, no panic. Where the evidence runs out, the book says so.