AI Superpowers

“When the Soviet Union launched the first human-made satellite into orbit in October 1957, it had an instant and profound effect on the American psyche and government policy. The event sparked widespread U.S. public anxiety about perceived Soviet technological superiority, with Americans following the satellite across the night sky and tuning in to Sputnik’s radio transmissions. It triggered the creation of the National Aeronautics and Space Administration (NASA), fueled major government subsidies for math and science education, and effectively launched the space race. That nationwide American mobilization bore fruit twelve years later when Neil Armstrong became the first person ever to set foot on the moon.”

This passage uses analogy to support Lee’s “Chinese AI victory” prediction. By dubbing Ke Jie’s match with AlphaGo a “sputnik moment,” Lee connects the Space Race to what he calls the “AI race” between the US and China. He argues that the USSR’s initial lead in the Space Race spurred the US to compensate and “win” by reaching the moon first. In the AI race, the US has the initial lead, and Lee predicts that China will be spurred to overtake them.

“Deep Blue had essentially ‘brute forced’ its way to victory—relying largely on hardware customized to rapidly generate and evaluate positions from each move. It had also required real-life chess champions to add guiding heuristics to the software. Yes, the win was an impressive feat of engineering, but it was based on long-established technology that worked only on very constrained sets of issues. Remove Deep Blue from the geometric simplicity of an eight-by-eight-square chessboard and it wouldn’t seem very intelligent at all. In the end, the only job it was threatening to take was that of the world chess champion.

Here, Lee explains why Deep Blue’s victories in the 1990s are not comparable to AlphaGo’s victory in 2016. This also demonstrates the rapid growth of AI’s capabilities over the course of roughly thirty years.

“Fundamentally, these algorithms use massive amounts of data from a specific domain to make a decision that optimizes for a desired outcome. It does this by training itself to recognize deeply buried patterns and correlations connecting the many data points to the desired outcome. This pattern-finding process is easier when the data is labeled with that desired outcome—”cat” versus “no cat”; “clicked” versus “didn’t click”; “won game” versus “lost game.” It can then draw on its extensive knowledge of these correlations—many of which are invisible or irrelevant to human observers—to make better decisions than a human could. Doing this requires massive amounts of relevant data, a strong algorithm, a narrow domain, and a concrete goal. If you’re short any one of these, things fall apart.”

This is a simplified explanation of deep-learning’s basic mechanisms. While this book is uninterested in explaining the technical side of AI’s development and implementation, this section serves as a quick crash course for laypersons.

“Deep-learning pioneer Andrew Ng has compared AI to Thomas Edison’s harnessing of electricity: a breakthrough technology on its own, and one that once harnessed can be applied to revolutionizing dozens of different industries. Just as nineteenth-century entrepreneurs soon began applying the electricity breakthrough to cooking food, lighting rooms, and powering industrial equipment, today’s AI entrepreneurs are doing the same with deep learning. Much of the difficult but abstract work of AI research has been done, and it’s now time for entrepreneurs to roll up their sleeves and get down to the dirty work of turning algorithms into sustainable businesses.”

Lee likens AI to a number of historical GPTs throughout this book. Here, he uses 19th-century breakthroughs in electrical engineering—arguably the most impactful scientific breakthrough of the 1800s—to emphasize the significance of AI research.

“The dramatic transformation that deep learning promises to bring to the global economy won’t be delivered by isolated researchers producing novel academic results in the elite computer science labs of MIT or Stanford. Instead, it will be delivered by down-to-earth, profit-hungry entrepreneurs teaming up with AI experts to bring the transformative power of deep learning to bear on real-world industries.”

This quote positions academic elitism and voracious capitalism as opposites. While Lee credits lofty American geniuses with creating AI, he believes that ambitious Chinese entrepreneurs will awaken this technology’s true practical potential.

“Silicon Valley may have found the copying undignified and the tactics unsavory. In many cases, it was. But it was precisely this widespread cloning—the onslaught of thousands of mimicking competitors—that forced companies to innovate.”

This quote demonstrates Lee’s disinterest in ethical and moral issues around the global tech economy. AI Superpowers is focused on convincing readers that China will “win” the AI race without commenting on that victory’s ramifications or virtues.

“Combine these three currents—a cultural acceptance of copying, a scarcity mentality, and the willingness to dive into any promising new industry—and you have the psychological foundations of China’s internet ecosystem.”

Lee takes a quasi-sociological approach to explaining Chinese tech culture. He largely pins its development on these general factors, presenting his western readership with a broad snapshot of China’s internet ecosystem.

“They see the primary job in China as marketing their existing products to Chinese users. In reality, they need to put in real work tailoring their products for Chinese users or building new products from the ground up to meet market demands.”

After emphasizing the cultural differences between Chinese and American tech, Lee argues that these differences (combined with the massive size of the US’s internationally expanding companies) pose a barrier for American expansion into the Chinese tech ecosystem.

“Foreign firms are often left with mild-mannered managers or career salespeople helicoptered in from other countries, people who are more concerned with protecting their salary and stock options than with truly fighting to win the Chinese market. Put those relatively cautious managers up against gladiatorial entrepreneurs who cut their teeth in China’s competitive coliseum, and it’s always the gladiators who will emerge victorious.”

This passage exemplifies Lee’s use of lyricism and literary craft. Throughout AI Superpowers, he uses well-turned phrases and elaborate analogies to convey his observations and predictions. Here, he conjures the image of a meek businessperson “fighting” a gladiator.

“Under the banner of ‘Mass Innovation and Mass Entrepreneurship,’ Chinese mayors flooded their cities with new innovation zones, incubators, and government-backed venture-capital funds, many of them modeled on Guo’s work with the Avenue of the Entrepreneurs. It was a campaign that analysts in the West dismissed as inefficient and misguided, but one that turbocharged the evolution of China’s alternate internet universe.”

The slogan “Mass Innovation and Mass Entrepreneurship” is a running motif in AI Superpowers. Its mere existence is a symbol of the Chinese tech ecosystem’s recent development. It also exemplifies Lee’s values succinctly.

“The four years leading up to Tencent’s Pearl Harbor moment saw many of the pieces of China’s alternate internet universe fall into place. Gladiatorial competition between China’s copycat startups had trained a generation of street-smart internet entrepreneurs. Smartphone users had more than doubled between 2009 and 2013, from 233 million to a whopping 500 million. Early-stage funds were fostering a new generation of startups building innovative mobile apps for this market. And WeChat demonstrated the power of the super-app installed on virtually everyone’s smartphone, an all-in-one portal to the Chinese mobile ecosystem.”

The title “Tencent’s Pearl Harbor moment” is another example of Lee’s propensity for historical comparisons.

“Those startups are now scrapping for a slice of an AI landscape increasingly dominated by a handful of major players: the so-called Seven Giants of the AI age, which include Google, Facebook, Amazon, Microsoft, Baidu, Alibaba, and Tencent. These corporate juggernauts are almost evenly split between the United States and China, and they’re making bold plays to dominate the AI economy. They’re using billions of dollars in cash and dizzying stockpiles of data to gobble up available AI talent. They’re also working to construct the ‘power grids’ for the AI age: privately controlled computing networks that distribute machine learning across the economy, with the corporate giants acting as ‘utilities.’ It’s a worrisome phenomenon for those who value an open AI ecosystem and also poses a potential stumbling block to China’s rise as an AI superpower.”

Here, Lee implies a distaste for the economic juggernauts of the AI age. The phrase “so-called Seven Giants” suggests doubts that these companies are and/or will continue to be “juggernauts.” In seeking to monopolize the infrastructure that makes AI possible, Lee argues that these companies are presenting an obstacle to innovation.

“In nuclear physics, the 1930s and 1940s were an age of fundamental breakthroughs, and when it came to making those breakthroughs, one Enrico Fermi was worth thousands of less brilliant physicists. American leadership in this era was built in large part on attracting geniuses like Fermi: men and women who could singlehandedly tip the scales of scientific power. But not every technological revolution follows this pattern. Often, once a fundamental breakthrough has been achieved, the center of gravity quickly shifts from a handful of elite researchers to an army of tinkerers—engineers with just enough expertise to apply the technology to different problems. […] A constant stream of headlines about the latest task tackled by AI gives us the mistaken sense that we are living through an age of discovery, a time when the Enrico Fermis of the world determine the balance of power. In reality, we are witnessing the application of one fundamental breakthrough—deep learning and related techniques—to many different problems. That’s a process that requires well-trained AI scientists, the tinkerers of this age.”

Lee’s framework of two mutually exclusive ages—ages of discovery and implementation, respectively—relies on the concept of genius to function. Lee takes for granted that 1) genius is a concrete, identifiable quality in certain individuals and 2) these individuals are able to “singlehandedly tip the scales of scientific power.”

“To some observers in the West, these research achievements fly in the face of deeply held beliefs about the nature of knowledge and research across political systems. Shouldn’t Chinese controls on the internet hobble the ability of Chinese researchers to break new ground globally? There are valid critiques of China’s system of governance, ones that weigh heavily on public debate and research in the social sciences. But when it comes to research in the hard sciences, these issues are not nearly as limiting as many outsiders presume. Artificial intelligence doesn’t touch on sensitive political questions, and China’s AI scientists are essentially as free as their American counterparts to construct cutting-edge algorithms or build profitable AI applications.”

In this passage, Lee’s ambivalence to ethical issues extends to political matters. He regards AI as fundamentally apolitical and only takes interest in Chinese governmental “controls” insofar that they impact the tech ecosystem’s economic viability.

“Smart Finance’s deep-learning algorithms don’t just look to the obvious metrics, like how much money is in your WeChat Wallet. Instead, it derives predictive power from data points that would seem irrelevant to a human loan officer. For instance, it considers the speed at which you typed in your date of birth, how much battery power is left on your phone, and thousands of other parameters. What does an applicant’s phone battery have to do with creditworthiness? This is the kind of question that can’t be answered in terms of simple cause and effect. But that’s not a sign of the limitations of AI. It’s a sign of the limitations of our own minds at recognizing correlations hidden within massive streams of data. By training its algorithms on millions of loans—many that got paid back and some that didn’t—Smart Finance has discovered thousands of weak features that are correlated to creditworthiness, even if those correlations can’t be explained in a simple way humans can understand.”

Here, Lee explains both the predictive power of weak features and the breadth of AI’s influence on China’s financial ecosystem. He also points toward some of the ethical quandaries raised by AI, as decisions that have large impacts on people’s lives are made by machines for reasons that can’t be explained or justified in human terms.

“Second-wave AI promises to change all of this. Underneath the many social elements of visiting a doctor, the crux of diagnosis involves collecting data (symptoms, medical history, environmental factors) and predicting the phenomena correlated with them (an illness). This act of seeking out various correlations and making predictions is exactly what deep learning excels at. Given enough training data—in this case, precise medical records—an AI-powered diagnostic tool could turn any medical professional into a super-diagnostician, a doctor with experience in tens of millions of cases, an uncanny ability to spot hidden correlations, and a perfect memory to boot.”

Here, Lee exemplifies the power of predictive AI. He also sets up a scenario which he will revisit in Chapter 8: the notion of AI taking over diagnosis in medical settings.

“That type of data collection may rub many Americans the wrong way. They don’t want Big Brother or corporate America to know too much about what they’re up to. But people in China are more accepting of having their faces, voices, and shopping choices captured and digitized. This is another example of the broader Chinese willingness to trade some degree of privacy for convenience.”

This is another instance of Lee attributing China’s chances of success with AI to broad cultural trends. Citizens who are more willing to tolerate surveillance produce more data.

“Getting to AGI would require a series of foundational scientific breakthroughs in artificial intelligence, a string of advances on the scale of, or greater than, deep learning. These breakthroughs would need to remove key constraints on the ‘narrow AI’ programs that we run today and empower them with a wide array of new abilities: multidomain learning; domain-independent learning; natural-language understanding; commonsense reasoning, planning, and learning from a small number of examples. Taking the next step to emotionally intelligent robots may require self-awareness, humor, love, empathy, and appreciation for beauty. These are the key hurdles that separate what AI does today—spotting correlations in data and making predictions—and artificial general intelligence. Any one of these new abilities may require multiple huge breakthroughs; AGI implies solving all of them.”

Here, Lee breaks down the low chances of the singularity occurring in our lifetimes (or possibly ever). The assertion that “humor, love, empathy, and appreciation for beauty” eludes AI lays the groundwork for his later assertion that humans can survive the impending AI jobs crisis by centering those uniquely human capacities.

“Unlike the GPTs of the first and second Industrial Revolutions, AI will not facilitate the deskilling of economic production. It won’t take advanced tasks done by a small number of people and break them down further for a larger number of low-skill workers to do. Instead, it will simply take over the execution of tasks that meet two criteria: they can be optimized using data, and they do not require social interaction.”

The Industrial Revolution made certain forms of skilled labor obsolete (for example: weaving). The positions those laborers held prior to industrialization disappeared and were replaced with low-skill jobs in factories (e.g., power loom operation). Lee points out that, while the invention of AI is similarly revolutionary, the way it will impact jobs today is not analogous to the way industrialization impacted them in the 18th and 19th centuries. AI has the potential to completely automate a slew of fields without creating new ones.

“As a technology and an industry, AI naturally gravitates toward monopolies. Its reliance on data for improvement creates a self-perpetuating cycle: better products lead to more users, those users lead to more data, and that data leads to even better products, and thus more users and data. Once a company has jumped out to an early lead, this kind of ongoing repeating cycle can turn that lead into an insurmountable barrier to entry for other firms.”

Although he is largely in favor of the “virtuous cycles” created by AI’s introduction to capitalist markets, Lee also recognizes that these cycles can hinder competition by magnifying the advantages of the companies that establish themselves first.

“Yes, this technology will both create enormous economic value and destroy an astounding number of jobs. If we remain trapped in a mindset that equates our economic value with our worth as human beings, this transition to the age of AI will devastate our societies and wreak havoc on our individual psychologies. But there is another path, an opportunity to use artificial intelligence to double down on what makes us truly human. This path won’t be easy, but I believe it represents our best hope of not just surviving in the age of AI but actually thriving. It’s a journey that I’ve taken in my own life, one that turned my focus from machines back to people, and from intelligence back to love.”

This quotation marks a shift in Lee’s core argument. While Chapters 1-5 focus on the economic ramifications of AI, Chapter 6 begins to examine the potential human costs of these successes. Chapters 7-9 prescribe solutions to these issues that center human compassion.

“Traditional doctors could instead evolve into a new profession, one that I’ll call a ‘compassionate caregiver.’ These medical professionals would combine the skills of a nurse, medical technician, social worker, and even psychologist[…] These compassionate caregivers wouldn’t compete with machines in their ability to memorize facts or optimize treatment regimens. In the long run, that’s a losing battle. Compassionate caregivers would be well trained, but in activities requiring more emotional intelligence, not as mere vessels for the canon of medical knowledge.”

Lee presents a speculative scenario for AI augmenting skilled positions rather than outright replacing them. He emphasizes doctors’ roles as caregivers above all other functions, which he believes AIs will soon be able to perform beyond human capacity.

“As a venture-capital investor, I see a particularly strong role for a new kind of impact investing. I foresee a venture ecosystem emerging that views the creation of humanistic service-sector jobs as a good in and of itself. It will steer money into human-focused service projects that can scale up and hire large numbers of people: lactation consultants for postnatal care, trained coaches for youth sports, gatherers of family oral histories, nature guides at national parks, or conversation partners for the elderly. Jobs like these can be meaningful on both a societal and personal level, and many of them have the potential to generate real revenue—just not the 10,000% returns that come from investing in a unicorn technology startup.”

Lee believes that venture capitalists’ methods can (and will) be repurposed for humanitarian efforts in the near future. He suggests that human compassion will supersede their desire for profits—a stark contrast to his description of entrepreneurial “gladiators” in Chapter 2.