RALEIGH – Long dismissed as a technology of the distant future, artificial intelligence was a project consigned to the fringes of the scientific community. Then two researchers changed everything.
In his new book, Genius Makers, Cade Metz traces how their ideas drove a new kind of arms race, spanning Google, Microsoft, Facebook, and OpenAI, a new lab founded by Silicon Valley kingpin Elon Musk.
On Tuesday, the Raleigh native and NY Times technology correspondent will hold a virtual launch at Quail Ridge Books. WRAL TechWire’s Chantal Allam had the chance to find out more. Here’s what he had to say:
- This week, you’re about to launch Genius Makers. Talk us through how this book came to be.
About four years ago, when I was working as a writer with Wired magazine, I traveled to Seoul, South Korea for an event that turned out to be far more interesting — and far more emotional — than anyone expected, including me.
A machine, built by researchers at Google, beat one of the world’s best players at the ancient game of Go, the Eastern version of chess. Go is so complex — requiring not just skill but what we humans call intuition — that even the experts thought it would be decades before a machine could beat the top humans.
The match in Korea was an inflection point for what we call “artificial intelligence.” When I got back to the States, I started work on the book.
- Through the lives of a small group of scientists, you explore the birth of artificial intelligence. Among them is Geoff Hinton, a man born into a family of eccentric scientists who became known as the father of deep learning. What is it about him and other key players that make for such a compelling narrative?
When I started outlining the book, I knew that one of the central figures would be Demis Hassabis, who led the Google team that went to Seoul. A former chess prodigy who studied as a neuroscientist before vowing to build a machine that could do anything the human brain could do, he is an usual mix of intellect, ambition, and a kind of charisma that I hadn’t seen before.
But as I built the book, what I discovered was a much larger cast of characters, each just as interesting and surprising and charmingly unusual as the last. In the end, the central character became Hinton, a London-born researcher who does not sit down (literally). He hurt his back as a teenager, and now, the only way to keep the pain at bay is to avoid sitting down, which means he can’t drive a car and he can’t fly on an airplane. This became a metaphor for the book’s central narrative. So much of what has happened over the past ten years with the rise of artificial intelligence — including digital assistants like Siri that can recognize what you say, face recognition on Facebook and other internet services, self-driving cars, and other robotics — is driven by a single mathematical idea. This is an idea that Hinton embraced in the early 1970s, when almost no one believed in it.
And despite enormous skepticism over the decades, he continued to believe it. In about 2010, this idea finally started to work, and at that moment, Hinton — and the few other academics who worked on this idea — were sucked into some of the biggest companies on Earth. Google. Facebook. Microsoft. Apple. Baidu in China.
- Your book is crammed with behind-the-scenes anecdotes about how AI entered Google, Facebook, and the rest of high tech. Which ones stand out to you the most?
The opening of the book, when Hinton auctions his services to the highest bidder, is certainly on the list. It is such an amazing story, the book just had to begin that way. But there are so many fascinating and shocking moments, just because the stakes become so high for these characters — and they are wonderfully free-thinking people. I am also partial to the moment in the book when Qi Lu, one of the top executives at Microsoft, decides he is going to change the company’s direction by riding a backwards bicycle. It goes right when you turn the handle bars left, and it goes left when you turn right. For the punchline, you’ll have to read the book …
- Google launched an in-house AI lab. Microsoft did the same. Then Elon Musk jumped in, founding OpenAI, stealing talent from both Google and Facebook. These companies began waging an unprecedented arms race – which now includes China and Russia. Where do you see this all headed?
Well, what I can say is that this race will continue to escalate for the foreseeable future. The stakes are enormous, including not just self-driving cars and internet services but autonomous weapons and surveillance technologies. China is a particularly important player. And the way we, as a country, will compete with China is very different from the way we competed with rivals in decades past. Just recently, Eric Schmidt, the former CEO of Google, delivered a report to Congress and the President calling for what you might call a West Point for AI researchers — a national academy that would train government personnel for a new future.
- You grew up in Raleigh, attended Broughton High School, and later graduated from Duke University. Meanwhile, your father was a programmer and engineer at IBM at RTP, and he worked on the barcode. Do you have any stories that you can share from that time?
My father, Walt Metz, helped test the Universal Product Code, the barcode on all your groceries, when he was working at IBM in the ‘70s. I grew up on stories about that project, which began with one of my father’s colleagues, a man named Joe Woodland. What people may or may not realize is that when the UPC was first deployed, many people protested outside grocery stores because they saw the barcode as “the sign of the beast” from the Book of Revelations. I once wrote a story about this.
AI is now a part of our everyday lives making everything from talking digital assistants and driverless cars to drug discovery possible. But as you present in your book, the vision and hopes guiding the founding scientists hasn’t always aligned with the realities of the world, money and greed. Talk to us about that, and what the implications are for the future.
This became the heart of the book. The problems and questions around these technologies are what drive the second half of the narrative. There are so many issues that we, as a society, need to think through. Many of these AI systems are biased against women and people of color. We now have technology that can write its own tweets, blog posts, speeches, and news articles. But this technology, because it learns from the way humans write, can also spew hate speech and other toxic content. The same idea that Hinton worked on — called a neural network — can also generate photo-realistic images and videos. We are entering a world where it will be very hard to tell whether what we are looking at on the internet is real or fake. Then there’s the problem of autonomous weapons and face recognition technologies. The list goes on.
- Why is this one of the most important stories of our time? And looking ahead, how far do you think we will let it go?
It is important because of all problems I just listed above and more — not to mention all the positives that these technologies will bring. Drug discovery is a very important area here. There are signs that AI technologies will allow us to tackle the next pandemic much quicker than we dealt with this one.