China's Awakening — AlphaGo and a National Plan

A Go board is nothing more than a plain wooden plank where nineteen vertical and horizontal lines cross. Yet the number of possibilities that unfold upon it is said to exceed the number of atoms in the universe. For that very reason, Go was believed to be the last fortress that machines could never breach — the one place where they would never surpass human beings. In the spring of 2016, the world watched in Seoul as that fortress fell. And across the sea, a great power received that single match as the starting gun for a change in its own destiny.

The Five-Game Match in Seoul

In March 2016, in a hotel room in Seoul, South Korea, a contest began that held the world’s breath. Seated across the board was Lee Sedol, a legendary master of the Go world. His opponent was not human. It was a program called AlphaGo, developed by the British research organization DeepMind.

As we saw last time, since the deep-learning revolution of 2012, AI had shown a power that rivaled humans in image recognition. But Go was considered to be in a class of its own. The choices were simply too many; memorizing standard openings alone could never be enough to win. It demanded intuition and a grasp of the whole board — or so it was thought. Many experts believed that a machine defeating a top human player was still a decade away.

Yet the match overturned that prediction with ease. In the best-of-five series held from March 9 to 15, AlphaGo prevailed four games to one. Lee Sedol struck back in the fourth game, but the machine had beaten the world champion. The number of people who watched this match around the world is said to have exceeded two hundred million.

A “Sputnik Moment” for China

This news was reported widely around the world. But one country received it in a way that was decisively different from the rest. That country was China.

Why did China react so strongly? The key to understanding it lies in a single phrase. Kai-Fu Lee, an AI researcher and investor, would later describe the event as China’s “Sputnik moment.”

Sputnik was the world’s first artificial satellite, launched by the Soviet Union in 1957. The United States, which had believed itself the world leader in science and technology, was deeply shaken to find itself overtaken in the space race, and made it a turning point for national investment in science education and space development. A single satellite rewrote the whole national strategy of a superpower.

According to Kai-Fu Lee, AlphaGo’s victory carried exactly that same meaning for China. An American-born AI had defeated the finest human player at Go, a game born in the East — and this spectacle lit a fire in the hearts of China’s government, its companies, and its young entrepreneurs. AI, they felt, was the technology that would decide the next era of supremacy. They could not afford to fall behind here.

The State Sounds the Call for AI

And so, the following year, in 2017, China set down that resolve in writing as the will of the state.

On July 20, 2017, China’s highest administrative body, the State Council, released a single document: the “New Generation Artificial Intelligence Development Plan.” It was a declaration that AI would be placed at the core of national strategy — a blueprint for the state, in effect.

Its goal was clear. Setting staged milestones for 2020 and 2025, it aimed ultimately to reach a world-leading level in AI theory, technology, and application by 2030, and to make China the world’s primary center of AI innovation. It envisioned growing the AI industry itself to the scale of one trillion yuan, and, counting related industries, an economic sphere ten times that size — a grand blueprint indeed.

The Rise of BAT

It was not only government agencies that began to run at the state’s call. China already possessed giant tech companies that had grown to be among the largest in the world. They were three firms, known by their initials as BAT.

The first was the search giant Baidu. Called “China’s Google,” it held the domestic market in search and advertising, and had been pouring effort into autonomous driving and AI research from an early stage. The second was the e-commerce titan Alibaba. With its vast trove of transaction data and its network of payments, it expanded its business into cloud computing and AI. The third was Tencent, known for the messaging app WeChat. Upon the data that some billion users exchanged each day, it built its AI services.

These three firms, riding the tailwind of the national plan, each became a standard-bearer for AI in its area of strength. Their advantage was not technological skill alone. A population of more than a billion searched, shopped, sent messages, and moved through the real world — and every one of those gestures flowed in as vast data to train AI. China’s AI stood upon a “fuel” supply unlike any other in the world.

In this way, in little more than a year, China built a system that placed AI at the core of its bid for global power. The awakening that began with a single match in Seoul transformed into concrete power: the partnership of a national plan and giant corporations.

Yet there is a core of this story still left untold. Why was it believed that China could climb to the rank of an AI great power in so short a time? The answer lay less in the technology itself than in the location of a certain “resource.” The new resource that decided victory lay not underground, but inside the smartphone. Next time: the story of “data,” the strategic commodity of the twenty-first century.