Opens in a new tab
16 min 02 sec
Available in:English

Genius Makers: The Mavericks Who Brought AI to Google, Facebook, and the World

Cade Metz tells how a few researchers turned a long-dismissed idea, machines that learn from examples, into the AI behind Google and Facebook, and why that same power brings fakes, bias and political risk.

Table of Content

Most of us use artificial intelligence every day without knowing where it came from, what it can really do or when to doubt it. In Genius Makers, New York Times technology reporter Cade Metz shows that today’s AI grew from one idea experts dismissed for decades: a machine that learns by finding patterns in huge piles of examples. A few stubborn researchers kept that idea, the neural network, alive through a long freeze, and then Silicon Valley’s giants paid fortunes for it.

Built on interviews with hundreds of specialists at Google, Microsoft, OpenAI and elsewhere, the book follows the technology from a 1958 flash-card demonstration to machines that beat champions, read eye scans, fake videos and serve armies. One learning method drives all of those outcomes, and whoever feeds and funds it decides where it points. Hold on to that thread and you can explain modern AI in two sentences, question a dazzling demo and see who is steering the technology.

A machine the size of a refrigerator learned to read flash cards in 1958, and then one influential critic froze the whole field for years. The trick it performed still sits underneath almost every AI you use.

A researcher nobody wanted to hire spent decades on an idea his field had abandoned. One conversation at a conference, plus some chips built for video games, changed what his work was worth.

A young researcher’s phone rang one evening, and the caller was one of the most famous chief executives in tech. The giants were spending millions on people before they knew how the work would ever pay.

An ancient board game held out against computers for decades, and a common eye disease needs specialists who are often in short supply. Both gave way to the same method, for reasons that have little to do with genius.

A president speaking flawless Mandarin, and face recognition that stumbles on anyone who is not a white man. Both problems grow from the same source, and knowing it changes how you judge what you see online.

Eight engineers thought they were building an ordinary image tool for a client. What they learned about that client, and what thousands of Google staff did next, shows who really steers this technology.

A phone call to a restaurant fooled the person who answered and thrilled a packed amphitheater. A psychology professor watching it was unmoved, and his reason has to do with how babies learn.

Genius Makers is a story about patience and about power. For roughly half a century, the idea that a machine could learn from examples was a minority view, kept alive by people like Rosenblatt and Hinton. When fast chips and endless data arrived, that single idea did almost everything at once. It beat a Go champion, read eye scans as well as a doctor, booked a restaurant table, put foreign words in a politician’s mouth and helped guide drones. In Metz’s telling these are not separate technologies with separate morals; they are one learning loop, pointed wherever its funders choose.

That gives you a way to hear the next AI headline. Ask what the system learned from and whose data was missing. Ask who is paying for it. And when a demonstration thrills you, remember that a baby grasps from two examples what a network needs millions to learn.

Rosenblatt’s flash cards remain the best picture to keep: a machine getting better at whatever it is shown, and someone always choosing the cards.

About this book

What is this book about?

Artificial intelligence now sorts photos, translates languages, aims ads and even answers the phone, yet most people have little idea where it came from or how far to trust it. Cade Metz, who covers AI for the New York Times, tells the story through the people who built it, drawing on hundreds of interviews with specialists at companies such as Google, Microsoft and OpenAI. The story begins in 1958 with Frank Rosenblatt's Perceptron, a machine that learned to read flash cards, and moves through the AI winter that followed Marvin Minsky's attack on the idea. It follows Geoff Hinton's decades as an outsider, his deep learning breakthrough in speech recognition with Microsoft's Li Deng, and the bidding war in which Google, Facebook and Microsoft spent millions on researchers. Metz then weighs what the technology has done: AlphaGo beating Go champions, a network spotting an eye disease about as well as a doctor, deep fake videos, biased face recognition, drone contracts and election scandals. Readers get a clear picture of how neural networks learn, who controls them and why critics such as Gary Marcus doubt they think the way we do.

Book Information

Rating:

Genre:

Technology & the Future

Topics:

Artificial Intelligence, Innovation

Publisher:

Penguin Random House

Language:

English

Narrator:

Jenna

First published:

March 16, 2021

Pages:

384

Length:

16 min 02 sec

Get the book:

Amazon

Who it's for

  • Professionals using AI tools

    Professionals who use AI tools at work and want to know how these systems actually learn.

  • Weighing AI investments

    Founders and managers weighing AI investments who want the history behind the hype.

  • Uneasy about AI's risks

    Curious readers uneasy about deep fakes, biased algorithms and AI in politics.

Listen to Preview

About the Author

Cade Metz

Cade Metz is a technology reporter at the New York Times, where he covers artificial intelligence, robotics and other newer digital technologies. Before that he was a senior staff writer at Wired magazine. His years of reporting on the technology industry shape Genius Makers, which tells the story of AI through the researchers and companies who built it.

Ratings & Reviews

Ratings at a glance

4.3

Overall score based on 98 ratings.

What people think

Listeners enjoy this book's extensive history of AI's creators and its thorough look at deep learning's trailblazers. The work is praised for being easy to read; one listener considers it essential for tech fans, while another characterizes it as an exciting exploration of AI's early innovators. Furthermore, listeners commend the way it simplifies intricate ideas, and one review notes that it "seamlessly blends technical brilliance with storytelling." Listeners also appreciate the perspectives on AI's impact and the depiction of the individuals at the center of the field.

Top reviews

Kavya

Cade Metz has written a definitive account of the architects of machine intelligence. This book doesn't just explain the code; it chronicles the humans who bet their entire careers on neural networks when the rest of the world thought they were chasing a pipe dream. Personally, I was captivated by the portrait of Geoffrey Hinton, a man who literally could not sit down due to a back injury while he was upending the entire tech industry. The narrative flows with the energy of a corporate thriller, especially when describing the bidding wars between Google and Baidu for top-tier talent. It captures that rare lightning-in-a-bottle moment where academic curiosity transformed into billion-dollar infrastructure. While some technical readers might want more math, the focus on personalities like the eccentric Marvin Minsky and the visionary Demis Hassabis makes it incredibly accessible. It's a gripping story about how a small group of contrarians eventually won the future.

Show more
Tang

As someone who usually finds tech history dry, I was shocked by how emotional this journey felt. Metz focuses heavily on the human side of the equation, specifically the struggles of pioneers like Hinton who faced decades of rejection. The book manages to blend technical milestones with tragic personal details, like the loss of Hinton’s wives to cancer, which fueled his drive for medical AI. It makes these 'geniuses' feel like real people rather than just distant figures in a lab. The way the author describes the transition from university hallways to the high-stakes boardrooms of Google and Facebook is masterfully done. To be fair, it’s a very male-centric story, but that reflects the unfortunate reality of the field during those decades. I came away with a much better understanding of why my phone can recognize my face and why that technology is so complicated to get right. Absolutely brilliant storytelling.

Show more
Watcharee

What a wild ride through the history of a technology we use every single day without thinking. I’ve been following AI for a while, but I never knew the level of drama involved in the creation of things like GANs or BERT. The 'GANfather' Ian Goodfellow and his journey to Google is a particularly interesting thread. Metz has a knack for finding the small, telling details that make a story pop, like the specific graphics cards needed to run these massive simulations. This isn't just a book about computers; it’s a book about obsession and the price of being right when everyone else says you're wrong. In my experience, most tech books are either too simple or too complex, but this one hits the sweet spot perfectly. It’s informative, entertaining, and just a little bit frightening when you think about where we are headed. Highly recommended for anyone who wants to understand the brains behind the bots.

Show more
Gabriel

Wow, I didn't realize the world of artificial intelligence was so full of giant egos and petty rivalries! This book is a page-turner that feels more like a biography of an idea than a dry history of software. Cade Metz manages to make you care about the researchers as much as the research itself. I loved the bits of humor sprinkled throughout, especially Geoff Hinton’s dry wit regarding his own academic failures. The book successfully explains why 'deep learning' was such a breakthrough and why companies were willing to spend $650 million on a startup like DeepMind before it even had a product. It’s a fascinating look at the intersection of extreme intelligence and extreme wealth. Even if you aren't a 'tech person,' you will find the stories of these pioneers and their struggles worth the price of admission. It’s easily one of the best non-fiction books I’ve read this year.

Show more
Samira

Ever wonder how a handful of researchers became worth hundreds of millions of dollars? This book pulls back the curtain on the talent wars that shaped our current digital reality. Metz does a fantastic job translating the 'AI Winter' and the subsequent spring into a story that feels personal rather than academic. Frankly, the sheer number of names can be a bit overwhelming at times, and I found myself wishing for a character list to keep track of everyone moving between Google, Facebook, and OpenAI. That being said, the writing is sharp and the pacing is excellent. I particularly enjoyed the sections on AlphaGo; the tension in those matches felt real even though I already knew the outcome. It's a solid, well-reported history that highlights both the brilliance and the massive egos of the men involved. Not a perfect book, but definitely a must-read for anyone curious about the origins of the tools we use every day.

Show more
Joseph

Finally got around to reading this one after seeing it on several 'Best of' lists, and it mostly lived up to the hype. The book explores the dramatic shift from symbolic AI to the deep learning models that now power everything from our phones to our cars. Truth is, I expected more focus on the ethics and the 'gorilla' incident, but Metz treats those as secondary beats to the main story of technical development. He captures the cutthroat nature of Silicon Valley brilliantly, showing how even the most idealistic researchers eventually got swallowed by corporate giants. The prose is clear and moves quickly, which is a feat given how dense this subject matter can be. I did feel the ending was a bit abrupt, leaving the future of AGI mostly to the imagination. Still, it’s a fascinating look at the 'genius makers' and the immense power they now wield over our society.

Show more
Skylar

The chapter on DeepMind’s AlphaGo victory against Lee Sedol reads like a high-stakes thriller, and it’s easily the highlight of the book. Metz excels at explaining why these milestones mattered, showing how a game of Go was actually a proxy for the future of general intelligence. Gotta say, the author’s ability to demystify neural networks without using a single line of code is impressive. He takes you from the early days of the Perceptron in the 50s all the way to the massive language models of today. My only real gripe is that the middle section gets a little bogged down in the 'he-said, she-said' of academic rivalries. Some of the ego clashes between Minsky and the younger generation felt a bit petty and dragged on longer than necessary. Regardless, it’s a very readable bit of science history that avoids the typical hyperbole found in AI books. It provides a much-needed reality check on what these machines can and cannot do.

Show more
Cee

Metz delivers a sophisticated look at the 'gold rush' of modern computing with a focus on the key players who made it happen. The book does a stellar job of explaining the transition from academic theory to corporate product. Not gonna lie, I was fascinated by the descriptions of the early neural network machines built from old airplane parts. It really highlights how far we’ve come in such a short amount of time. The author balances the optimism of the researchers with the cold reality of how these tools are being used by the military and for surveillance. While the narrative is mostly pro-AI, it doesn't completely shy away from the 'GIGO' problem—garbage in, garbage out—and how human bias gets baked into the code. The prose is punchy and the research is clearly exhaustive. It’s an important book that helps ground the current AI hype in actual historical context.

Show more
Arjun

The Genius Makers provides a necessary bridge between science fiction and the reality of neural networks. Instead of worrying about Skynet, Metz shows us the real-world problems, like how a Google app misidentified a person as a gorilla. These are the human errors that define the field. The book is remarkably easy to read, avoiding jargon in favor of clear storytelling and vivid character sketches. I found the sections on the 'AI Winter' particularly enlightening, as they explain why progress stalled for so many decades. To be fair, the timeline jumps around a bit, which might confuse some readers, but the overall message is clear. It’s a story of perseverance and the eventual triumph of an idea that was once ridiculed. It leaves you with a profound sense of how much our lives have already been changed by these 'geniuses' and their machines. A solid four-star read for anyone interested in the future of humanity.

Show more
Somboon

Not exactly what I expected given the massive hype surrounding this title. While Cade Metz is clearly a talented journalist with incredible access, the book felt a bit like a collection of Wired articles stitched together. Look, the history is comprehensive, but the focus is almost entirely on a very small, homogeneous group of men. I was disappointed that the critical issues of bias and the social implications of facial recognition were treated as minor footnotes rather than central themes. The writing is polished, but the narrative arc feels repetitive after the third or fourth corporate acquisition. It’s a decent primer if you want to know which professor moved to which tech giant, but it lacks the critical edge I was hoping for. It’s more of a celebration than a deep investigation into the consequences of these technologies. Good for a plane ride, but don't expect a balanced critique of the industry’s darker side.

Show more
Show all reviews

Questions about this book

It tells the story of modern artificial intelligence through the people who built it. Metz follows neural networks, machines that learn by finding patterns in data, from Frank Rosenblatt’s 1958 Perceptron through a long AI winter to the deep learning boom, when Google, Facebook and Microsoft spent millions on researchers. He then weighs the results, from AlphaGo and medical diagnosis to deep fakes, bias and military contracts.

Hinton is the book’s central outsider. He earned his PhD at the University of Edinburgh in the early 1970s, when neural networks were out of favor, and kept working on them for decades. His idea of adding more layers, which he called deep learning, proved itself in 2009 when he and Microsoft’s Li Deng built a highly accurate speech recognition program. Google later bought his firm, DNNresearch.

Metz covers several. Generative adversarial networks, invented by Ian Goodfellow in 2014, made deep fake videos easy to produce. Joy Buolamwini showed in 2018 that face recognition from Google and Facebook worked worse on faces that were not white and male. AI also drew in the military, through a Clarifai drone project and Google’s Project Maven, and Facebook struggled with misinformation after the Cambridge Analytica scandal.

Not according to critics like Gary Marcus, whom Metz gives a full hearing. Marcus, a nativist, argues that much of human intelligence is built in by evolution, so a baby can recognize an animal from one or two examples while a neural network needs millions. Others, such as OpenAI’s Ilya Sutskever, believe humanlike general intelligence is reachable, and OpenAI made it an explicit goal in 2018.

The book suggests a few questions to ask of any AI system. What data did it learn from, and whose faces or voices were missing? Who is paying for it and for what purpose, as the Clarifai engineers learned to ask? And how narrow is the task behind an impressive demo? Replaying a striking video and looking for small glitches is another habit the deep fake chapter supports.

AUDIO SUMMARY AVAILABLE

Listen to Genius Makers in 15 minutes

Get the key ideas from Genius Makers by Cade Metz, plus 5,000+ more nonfiction titles.

  • 5,000+ titles
  • Listen as much as you want
  • 7-day free trial
  • Cancel anytime

Genius Makers

Preview · Full summary 16 min 02 sec
Home

Discover

Open app

Log in