The Four Waves of AI

This series of blogs, adapted from my book, AI Superpowers: China, Silicon Valley, and the New World Order, assesses the four waves of artificial intelligence: internet AI, business AI, perception AI, and autonomous AI.

The year 2017 marked the first time I heard Donald Trump speak fluent Chinese. During the U.S. president’s first trip to China, he showed up on a big screen to welcome attendees at a major tech conference. He began his speech in English and then abruptly switched languages.

“AI is changing the world,” he said, speaking in flawless Chinese but with typical Trump bluster. “And iFlyTek is really fantastic.”

President Trump cannot, of course, speak Chinese. But AI is indeed changing the world, and Chinese companies like iFlyTek are leading the way. By training its algorithms on large data samples of President Trump’s speeches, iFlyTek created a near-perfect digital model of his voice: intonation, pitch, and pattern of speech. It then recalibrated that vocal model for Mandarin Chinese, showing the world what Donald Trump might sound like if he grew up in a village outside Beijing. The movement of lips wasn’t precisely synced to the Chinese words, but it was close enough to fool a casual viewer at first glance. President Obama got the same treatment from iFlyTek: a video of a real press conference but with his professorial style converted to perfect Mandarin.

“With the help of iFlyTek, I’ve learned Chinese,” Obama intoned to the White House press corps. “I think my Chinese is better than Trump’s. What do all of you think?”

iFlyTek might say the same to its own competitors. The Chinese company has racked up victories at a series of prestigious international AI competitions for speech recognition, speech synthesis, image recognition, and machine translation. Even in the company’s “second language” of English, iFlyTek often beats teams from Google, DeepMind, Facebook, and IBM Watson in natural-language processing—that is, the ability of AI to decipher overall meaning rather than just words.

This success didn’t come overnight. Back in 1999, when I started Microsoft Research Asia, my top-choice recruit was a brilliant young Ph.D. named Liu Qingfeng. He had been one of the students I saw filing out of the dorms to study under streetlights after my lecture in Hefei. Liu was both hardworking and creative in tackling research questions; he was one of China’s most promising young researchers. But when we asked him to accept our scholarship offer and become a Microsoft intern and then an employee, he declined. He wanted to start his own AI speech company. I told him that he was a great young researcher but that China lagged too far behind American speech-recognition giants like Nuance, and there were fewer customers in China for this technology. To his credit, Liu ignored that advice and poured himself into building iFlyTek. Nearly twenty years and dozens of AI competition awards later, iFlyTek has far surpassed Nuance in capabilities and market cap, becoming the most valuable AI speech company in the world.

Combining iFlyTek’s cutting-edge capabilities in speech recognition, translation, and synthesis will yield transformative AI products, including simultaneous translation earpieces that instantly convert your words and voice into any language. It’s the kind of product that will soon revolutionize international travel, business, and culture, and unlock vast new stores of time, productivity, and creativity in the process.

The Waves

But it won’t happen all at once. The complete AI revolution will take a little time and will ultimately wash over us in a series of four waves: internet AI, business AI, perception AI, and autonomous AI. Each of these waves harnesses AI’s power in a different way, disrupting different sectors and weaving artificial intelligence deeper into the fabric of our daily lives.

The first two waves—internet AI and business AI—are already all around us, reshaping our digital and financial worlds in ways we can barely register. They are tightening internet companies’ grip on our attention, replacing paralegals with algorithms, trading stocks, and diagnosing illnesses.

Perception AI is now digitizing our physical world, learning to recognize our faces, understand our requests, and “see” the world around us. This wave promises to revolutionize how we experience and interact with our world, blurring the lines between the digital and physical worlds. Autonomous AI will come last but will have the deepest impact on our lives. As self-driving cars take to the streets, autonomous drones take to the skies, and intelligent robots take over factories, they will transform everything from organic farming to highway driving and fast food.

These four waves all feed off different kinds of data, and each one presents a unique opportunity for the United States or China to seize the lead. We’ll see that China is in a strong position to lead or co-lead in internet AI and perception AI, and will likely soon catch up with the United States in autonomous AI. Currently, business AI remains the only arena in which the United States maintains clear leadership.

Competition, however, won’t play out in just these two countries. AI-driven services that are pioneered in the United States and China will then proliferate across billions of users around the globe, many of them in developing countries. Companies like Uber, Didi, Alibaba, and Amazon are already fiercely competing for these developing markets but adopting very different strategies. While Silicon Valley juggernauts are trying to conquer each new market with their own products, China’s internet companies are instead investing in these countries’ scrappy local startups as they try to fight off U.S. domination. It’s a competition that’s just getting started, and one that will have profound implications for the global economic landscape of the twenty-first century.

Posted by Dr. Kai-Fu Lee on Jan 15, 2019 in

AI is Going to Change Everything

Dr. Lee Shares the 4 Coming Advances in AI

Download the Free White Paper