Scaling AI peaks one after another
On May 16, via a video link, United States President Donald Trump “addressed” a conference in Tianjin from Washington and floored the audience with his almost flawless Chinese.
Trump highlighted the big leaps made in artificial intelligence, or AI, but what impressed the audience more was the US president’s tone — his Chinese intonations, inflections and pitch were near perfect.
Well, as it transpired, the voice was not really Trump’s, after all, but that of an AI-enabled voice technology developed by iFlytek Co Ltd. And, for the record, unlike his granddaughter, Trump knows hardly any Chinese.
The iFlytek technology demonstrated its speech synthesis capability — it can produce an unbelievably human-like voice. The Trump video clip showcased iFlytek’s broad efforts to tap into voice computing, which is said to be the next major medium for man-machine interaction.
The man who helms the Anhuibased company is Liu Qingfeng, 45, a scientist-turned entrepreneur. Liu founded iFlytek in 1999, when he was a second-year PhD student at the University of Science and Technology of China.
In 2017, MIT Technology Review, an established science magazine published by the Massachusetts Institute of Technology, released a report naming iFlytek as the smartest Chinese company. To earn the accolade, it beat well-known contenders such as Tencent Holdings Ltd, Alibaba Group Holding Ltd and Baidu Inc.
At the global level, iFlytek ranks sixth, just below fifth-ranked Alphabet Inc. Companies are selected for the honor based on their achievements in combining innovative technology with an effective business model.
iFlytek has made a name for itself with products and services like voice-based digital assistants, realtime portable translators, automated court clerks that review certain types of cases, medical robots and AI-enabled assistants to primary school teachers, with a market footprint spanning 31 provinces across China and R&D laboratories in the United States.
The company’s 2017 profit was 2.8 billion yuan ($410 million) on a revenue of 5.45 billion yuan, with the commercial application of AI still in its infancy.
Liu’s mission to take voice recognition technologies beyond the experimental stage has been a long one. Almost no one believed he could succeed. In the late 1990s, the China market for voice recognition technologies was dominated by US companies such as IBM and Microsoft Corp.
Liu and his colleagues seemed to have nothing but a passion to learn and achieve something.
Then 26, Liu got a few of his university friends together and raised a seed fund from the university to found iFlytek. Soon, he realized being a scientist and an entrepreneur are two fundamentally different things.
“When I was engaged in scientific research, I would bury myself in books and experiments, ignoring everybody else. But as an entrepreneur, I can’t be so capricious,” Liu says. “When meeting someone better than me, I can’t just think of beating him or her. Instead, if they are the best people I could recruit, I must be kind to them, encourage them and help them grow.”
The seasoned executive now often refers to himself as a “mother” who is on a night journey with a string of children. “The road ahead may be fraught with traps, pitfalls and people who want to cheat you. While walking on this unknown road, I must protect the kids following me.”
iFlytek realized the voice computing sector was set for explosive growth, thanks to breakthroughs in machine learning algorithms. Even as it made headway, BAT, or Baidu, Alibaba and Tencent, the triumvirate of Chinese internetbased businesses, forayed into the segment, presenting stiff competition in voice computing and strategic voice interaction.
“iFlytek’s biggest edge lies in our 19 years of consistent input into the research and development of original technologies. No matter how innovative others are branding their technologies, they must put them in commercial applications,” Liu says.
Currently, iFlytek accounts for 70 percent of China’s market in voice-based technologies, according to data from the Speech Industry Alliance of China. Its voice assistant technology is the Siri of China, and its real-time portable translator puts AI to remarkable use, overcoming dialect, slang and background noise to translate between Chinese and 33 other languages with high accuracy.
The company is also leveraging its technology repertoire to branch out into other AI-enabled sectors. It is developing an AI-enabled system to assist courts in reviewing four types of cases, namely murder, theft, telecom fraud and illegal fundraising.
Its medical robot also passed the written test of China’s national medical licensing examination in November last year. Now, it is being applied to hospitals in Anhui province to function as a general practitioner and help doctors treat diseases.
Its AI-system has also been applied in the classrooms of primary and middle schools across the country to help teachers better educate students.