Are we seeing the creation of a parallel universe for US and Chinese tech industries?
I think the answer is yes. In the past, US has dominated the world in technologies from P.C. operating systems, semiconductors, to servers, and even Internet. But ever since the rise of mobile technologies, China has really leveraged the large market with a huge amount of data and now is beginning to innovate and build great mobile apps on which there's a large amount of data being collected.
Therefore, fueling great development of AI technologies. We're in a stage where US and China are roughly equal in mobile and AI technologies. And I would project in the future, US will continue to dominate US developed markets, English speaking countries, Japan, and a few other countries. But China will rise, with Chinese software being used by increasingly more countries in other parts of the world.
Did Silicon Valley underestimate Chinese tech companies?
Absolutely. Ten years ago, it was reasonable because most Chinese companies at the time were copycats. They were using Silicon Valley ideas and building companies in China. So, for every Google, Facebook, Amazon, there was a Chinese equivalent.
However, in the past 10 years, a miracle happened. China's market is huge, four times larger than the US. Lots of venture capital money poured in. And the Chinese entrepreneurs are hardworking, tenacious, very good at iterating companies. And actually, using a winner-take-all, highly competitive environment, came up with business model innovations that in many cases surpass the Silicon Valley.
While the Silicon Valley is still better in breakthrough innovation, China actually has done an amazing job in execution, is generating an equal amount of value, and all of this data generated in this large market, large apps, is now becoming fuel to make Chinese AI improving rapidly and quickly. Also catching up with the United States.
How do I provide the right education to my children in the AI era?
I think the main thing happening in the AI era is routine jobs will increasingly be done by AI and people have to move on to things that people are really good at. And I'll give you two examples: In school, children study a lot of verbal and quantitative - that's what we are tested on the S.A.T. - but verbal needs to evolve from just being able to spell and put together sentences. AI can do that. It needs to evolve towards communication, connection, trust. This are things only people can do.
And then quantitative needs to evolve from arithmetic and doing math, all the way up to analysis, strategy, planning, curiosity, critical thinking and creativity. So, these are the things that we need to evolve our education so that our children can have skill set that cannot be replicated by AI. And those, after all, are why we are human. And hopefully that will bring education in the right direction.
Will the AI revolution create or destroy jobs?
The answer is yes. It will create and destroy jobs. Which will be more? Historically looking at technology revolutions - technology revolutions generally create more jobs than they destroy. So, it's likely AI will be the same way. The challenge, though, is that AI, when it destroys jobs, will do so completely. So routine jobs such as back office jobs and assembly line jobs and in the future, jobs like drivers and many office workers that do routine jobs, will be outright replaced by AI. So that's a significant issue.
AI will also create jobs, but we don't yet know what those are, when they will be. One thing that is known is that the jobs created by AI will be non-routine jobs. So, there is a big issue about routine job workers who lose their jobs and now have to be retrained for the non-routine jobs.
True or false: Superintelligent machines will one day rule over humans?
Very unlikely. The reason is today's artificial intelligence is also known as artificial narrow intelligence, which means the A.I. we have today deployed in Internet engines, all the way up to autonomous vehicles, are nothing more than fancy pattern recognition tools that we humans fully control. And, in order to evolve from these tools, which we control, to tools equally intelligent or even more intelligent than us, there will need to be at least 10 or 20 huge breakthroughs. Breakthroughs like the ability to create self-awareness have self-awareness and have compassion and ability to reason and plan and also what desire to rule over humans which doesn't exist in tools. So this is very unlikely. If it happens it will be in the very long term. We currently have no engineering path to getting there so we better focus on some real problems caused by A.I. such as security, privacy and inequality. So don't worry about it.
Will A.I. reduce poverty and inequality or make them worse?
On the positive side, A.I. is enabling all technologies and industries and will create a huge amount of wealth in the world measured by PwC as 16 trillion dollars in the next eleven years and more wealth ought to reduce poverty and inequality. Another aspect is that A.I. can be applied to health care and reduce the cost of health care and can be applied to education and make it more accessible. So that should also improve. However, A.I. has the following problems. First, much of the wealth generated by A.I. will go to the powerful Internet companies and big data companies like Google, and Facebook, and Tencent, and Alibaba. So they will get richer and A.I. will displace a lot of jobs, so poor people might lose their jobs. And A.I. will make lots of money for AI Superpowers like US and China, and may make other countries worse off. So the answer is we don't know but it's something we have to think about.