Since his bestselling book “AI Superpowers: China, Silicon Valley, and the New World Order” was published in 2018, Dr Kai-Fu Lee’s key predictions have already been realised: that artificial intelligence would become mainstream and that China would become a global leader in AI.
Speaking at a Chamber webinar on 15 July, Dr Lee explained that because AI depends on data, China’s huge population and fast-growing internet use are a major advantage.
“The company or country that will benefit the most from AI is the company or country that has the most data,” he said.
China’s dominance in the sector was a controversial view when the book was first published, but has now become widely accepted.
On the likely development of AI over the next few decades, Lee explained that 2019 saw a major breakthrough when a computer’s ability to understand language exceeded human capabilities for the first time. This ability to digest and analyse huge amounts of information will affect all business sectors.
Investors can use AI to create different hypothetical scenarios and be alerted about potential opportunities, while AI can also carry out research that previously a junior employee would need to take time to do. In advertising, meanwhile, AI enables individually targeted advertisements.
Autonomous driving is an area where AI is having a notable impact. Lee expects fully autonomous vehicles to be a reality within 15 to 40 years, but there will be challenges ahead.
For example, Tesla’s “Smart Summon” feature, used in parking lots, had some glitches when it was launched. But within two weeks the technology had already learned from all the new data it had gathered, and the updated version was a huge improvement.
“This is the key point of AI,” he said. “You launch it, it gets more data, and gets better over time.”
Lee predicts that autonomous driving will follow a gradual path starting with simple vehicles in confined spaces, such as forklifts. These will get more sophisticated over time, such as baggage carriers in airports, then trucks on highways, and taxis driving the same route from hotel to airport. At all of these stages more data will be gathered until eventually the technology will reach the Level 5 fully autonomous stage.
Healthcare is another sector where there is tremendous scope for progress, with AI combining our health records as well as information about our genes.
AI could also lead to a revolution in drug discovery. Currently, developing a new drug costs around US$2 billion, which means that pharmaceutical companies have little incentive to research rare diseases. But in the next 10 to 20 years, Lee foresees a reduction in costs of as much as 90%.
“Many rare diseases will become treatable,” he said, “and common diseases will have multiple treatments available based on your individual needs.”
Moving on to China’s economy, Lee said that he expects the nation will remain the world’s factory. Although wages have risen considerably in China, automation will make manufacturing more cost effective.
Does this mean that automation will replace people completely?
“The area of routine, repetitive manual work will inevitably be decimated by automation, and countries need to deal with that,” Lee said. But AI cannot do everything yet.
Automation can replace visual inspection and transport within factories, but cannot yet replace human dexterity in handling small, irregular or soft shapes. So workers who assemble mobile phones, for example, are quite secure, as they handle very small components. Lab technicians and radiologists, however, should be concerned as their jobs are more easily replaced by AI.
Ultimately, AI could make food and energy cheaper and more plentiful, and help solve global poverty – a scenario that is possible so long as the “haves” are willing to share with the “have nots” around the world.