The key ideas

AI 2041: Ten Visions for Our Future

Kai-Fu Lee, Chen Qiufan, 2021

An insurance app that decides who you're allowed to date. A video you can't trust. A job taken over by an algorithm. AI expert Kai-Fu Lee and science fiction writer Chen Qiufan show the world of 2041 through ten stories from different countries, and after each one they honestly explain the technology behind it, how it will help and where its risks lie.

Who it is for

For anyone who wants to understand how artificial intelligence will change work, health, relationships and money over the next twenty years — without formulas and without panic.

The future with AI isn't predetermined: it will be shaped by the goals we give our machines.

What the book is about

Picture an ordinary day in 2041. Your insurance company's app gently hints at who you'd better not date. A robot brings your groceries. And a video of a famous politician saying something monstrous may turn out to be a fake that no human could spot with the naked eye. Sounds like a TV series? Kai-Fu Lee, one of the best-known artificial intelligence experts in the world and the former head of Google China, is convinced: it's not a TV series but a fairly sober forecast. And that future is less than twenty years away.

Together with science fiction writer Chen Qiufan, he came up with an unusual book. Chen writes ten stories about people from different countries — India, Nigeria, Korea, China, Japan, Sri Lanka, Iceland, the US, the Middle East and Australia. After each story, Kai-Fu Lee calmly breaks down the technology the plot rests on: how it works, what already works today, how it will help and how it could do harm. The result is a guidebook to the near future with both a heart and a cool head. And the fact that you're opening a book like this instead of brushing the topic off with “that's for programmers” is already a brave step: many people prefer not to think about the future until it knocks on the door.

Idea 1. Science fiction without fantasy

Why is it that when we hear “artificial intelligence,” the Terminator pops into our heads? Kai-Fu Lee believes that this very image keeps us from seeing the real changes. Movies taught us to fear a superintelligence that will rise up against humans, while real AI changes our lives quietly: through the news feed, credit scoring, recommendations, medical scans.

That's why the authors set themselves a strict rule: the stories include only technologies that, by their estimate, are highly likely to become reality by 2041. No magic and no “thinking machines with feelings.” The mechanism is simple: a story makes you live through a technology from the inside, feel how it changes relationships, work, family. And the analysis after the story brings you back down to earth and explains what's true and what's artistic license.

The stories span the whole world — from India and Nigeria to Iceland and Australia. That's no accident: AI changes life everywhere, but in different ways — depending on culture, laws and who holds the data.

What this means for you: the main question isn't “will robots take over the world” but “how are the algorithms that already exist changing my decisions, and what am I paying for that convenience.”

You scroll through news about “dangerous AI” and picture an army of metal robots. Meanwhile, your banking app has already decided what credit limit to give you, and your social feed has decided what will make you angry today. Not a single robot in sight, yet AI is already part of your day.

The thing to fear isn't an imaginary superintelligence but the invisible algorithms that are already here.

Read in Livria

About the author

Kai-Fu Lee is one of the world's leading experts in artificial intelligence, the former head of Google China and an investor. Chen Qiufan is a Chinese science fiction writer and the author of the novel “Waste Tide.”

All books by this author: Kai-Fu Lee, Chen Qiufan

Terms from the book

  • Deep learning
  • Brain-computer interface
  • Objective function: What an AI tries to maximize: profit, clicks, time spent in an app. The machine pursues this goal literally.
  • Deepfake: A fake video, photo or voice created by a neural network so well that it's hard to tell from the real thing.
  • GAN: Generative adversarial network: two neural networks, one creating a fake and the other trying to expose it, with each learning from the other.
  • XR: An umbrella term for virtual, augmented and mixed reality — technologies that blend the digital and real worlds.
  • Qubit: The unit of information in a quantum computer. Unlike an ordinary bit, it can be in a state that combines 0 and 1.
  • Federated learning: A way of training AI directly on people's devices without collecting their personal data on a single server.

All terms

Similar books