As a daughter, mother, Chinese immigrant to the United States, and leading artificial intelligence researcher, Fei-Fei Li has seen many different worlds. Her life experience also inspired her academic expertise—training computers to recognize and interpret images.
As Li writes, it’s surprising how the human brain—”a wet, organic hunk about five inches in diameter”—can outperform warehouse-sized supercomputers when it comes to understanding the “essence” of everything we see.
Show a photograph of a Thanksgiving dinner in front of any American and most could immediately guess the relationships between the family members and realize that they were eating turkey.
How does this conceptual magic work? In some ways, this instinctive ability to understand context defines what it is to be human. “To understand how we see, therefore, is to understand ourselves,” she writes.
In “The Worlds I See,” Li explores three themes with skill and thrilling insight.
It provides a clear and accessible history of machine learning, describing how this conceptual transformation occurred. Li also makes a heartfelt plea to keep the human side at the center of AI, to ensure we maximize the benefits of our new technological powers.
But most of all, she tells a captivating story about how a poor Chinese immigrant girl could arrive in the US and end up at the forefront of a scientific revolution. The book is a loving portrait of Li’s family past as well as an exploration of our technological future in an expert connection of the personal and professional.
“I was born an only child to a family in a state of quiet turmoil,” Li writes of her childhood in the 1980s in Chengdu, capital of Sichuan province. Like many of China’s ancient cities, Chengdu’s narrow streets, open courtyards, and green landscapes were being obliterated by soulless Soviet architecture.
But this bleak urban backdrop contrasted with the warmth and imagination demonstrated by her parents, united in ways no one else could understand. “Both had lost faith in the institutions of the world, and this made them partners, even accomplices, in a daily practice of defiance,” Li writes.
His father, whose job in the computer department of a chemical company was more of a façade than a career, had a pathological allergy to seriousness, an endless curiosity about nature, and a goofy ingenuity.
He came to life when he searched for butterflies, watched water buffalo frolicking in the rice fields, and captured rodents and insects to raise as pets.
And although he missed the birth of his daughter because he was distracted on a birding trip, he whimsically named her Fei-Fei. In Mandarin, “fei” means to fly.
When written in a Chinese character, his name resembles a bird. “The greatest compliment I can pay my father is also the most damning criticism: that he is exactly what would result if a child could project his ideal father in the total absence of adult supervision,” Li writes.
Li’s mother, a high school teacher turned office worker, was also a unique person, “an intellectual trapped in a life of imposed mediocrity.” Her own grandmother had been one of the first women to attend university.
But his mother’s intellectual career was hampered by the family’s associations with the defeated Kuomintang party in opposition. Her life was also plagued by health problems, which sapped even her fierce energy.
Her coping mechanism was to escape into literature, a passion she passed on to her daughter, who devoured Chinese translations of Simone de Beauvoir, Charles Dickens and Ernest Hemingway.
At school, Li was outraged to hear a teacher tell boys that they were “biologically smarter than girls.” This silent heat of indignation fueled her studies as she fell in love with the beauty and grandeur of physics.
Li’s father moved to the United States to start a new life for his family. After three years apart, Li, then 15, and his mother joined him in New Jersey.
Li barely spoke a word of English when she enrolled at Parsippany High School, where everything was “brighter, faster, heavier and louder” than the world she had left behind. But she was lucky enough to be taught by Bob Sabella, or the “big bearded math teacher” as her parents called him.
Thanks to her inspiring guidance — and financial support — Li earned a place at Princeton University, a world apart from the laundry room her parents ran.
Li leaned into the field of computer vision, a relatively neglected area that was being transformed by the deep learning revolution.
His main contribution to academic fame was his role in creating ImageNet, which, at the time of its launch in 2009, was the largest manually curated dataset in the history of AI, containing 15 million images in 22,000 categories, annotated by more of 48,000 employees from 167 countries.
Combined with the application of smarter algorithms and massive computing power, ImageNet has triggered explosive growth in computer vision capability. These advances helped researchers develop self-driving cars and text-imaging models like Stable Diffusion, Dall-E, and Midjourney that impress users so much today.
The third theme of the book is a discussion of how we can maximize the benefits of AI while minimizing its potential harms.
As Li explains, a subject that was once the isolated domain of expert scientists must now concern everyone. After decades spent “in vitro”, AI is now very much present “in vivo”, as she puts it.
Technology researchers and world leaders mingling at the AI security conference in Bletchley Park this week show how the topic has gained ground on the global political agenda.
As one of the few female leaders in the field, Li is particularly concerned about the alarming lack of diversity in the industry. And having worked as chief AI scientist at Google Cloud, she says she is disturbed by the vast wealth, power and ambition of big tech companies, which dwarf anything in the university sector.
In 2015, for example, Uber hired almost the entire robotics team at Carnegie Mellon University to develop its autonomous cars.
To redress the imbalance, Li helped launch AI4All (AI for All) to encourage more diverse people to enter the industry. She also co-founded the Stanford Institute for Human-Centred Artificial Intelligence, which aims to strengthen AI research, education, and policy for the public good.
“The Worlds I See” ends with a sense of wonder at our ability to create a technology akin to a force of nature, “something so big, so powerful and so capricious that it could destroy as much as it could inspire.” But, she concludes, , it will take much more than corporate clichés to make AI worthy of our trust.