In the past 15 years, the advance of technology has affected all of our lives. The adoption of technology has been so vast and deep that our grandparents and children are all using some type of “smart” product every day.
In many ways, what we previously saw only in sci-fi movies is becoming our reality. Software and programmes can learn and become smarter. Hardware and machines can interact with each other and improve. Artificial intelligence (AI) technology, which I researched academically in the 1980s, is penetrating the real world and permeating business operations and individuals’ lifestyles without us even noticing.
Think about the last time you: received a recommendation about a cool new product on an e-commerce website; had a photo taken of you at an immigration office; or received a customer service chat or email reply after raising a complaint.
Buzz around AlphaGo was a wake-up call to the world. With decades of lab research, AI technology is finally coming to its anticipated fruition. I estimate the AI market potential will be 10 times greater than that of the mobile internet revolution we have witnessed thus far.
The time is right to open Pandora’s box for laboratory AI, introducing it to solve real business needs and address real-world problems. It is also the time for some scientists in the field to consider starting their entrepreneurial journey.
In the next decade, more than 50% of jobs in the world will be replaced by AI, ranging from translators, editors, assistants, stock traders, securities, drivers, salespeople, customer service reps, accountants, nannies and so on.
The five building blocks of AI development
AI development requires five building blocks: massive data, automatic data tagging systems, top scientists, defined industry requirements and highly efficient computing power. All of these are now tangible and available, because of the rise of the internet, mobile, big data and computing processing capabilities in the past 10 years.
So what are the action steps ahead? To start, we need to identify industries with an enormous amount of data. Better yet, the data itself should be as closed-loop and as massive as possible.
Then we need lots of computers, especially those with a combination of a high performance central processing unit (CPU) and high graphics processing unit (GPU).
Most importantly, we need great scientists specialized in deep learning, coupled with a group of young engineers eager to learn, experiment and solve problems. I am confident that graduates from the top universities for computer sciences, mathematics, applied mathematics, statistics, electronics or automation, are trained with the essentials to further develop AI skill sets. With guided training, they would be able to perform within six to nine months.
I would suggest a few guiding principles for the fast-growing commercial adoption of AI. Firstly, we should see AI as a tool to assist humans, not to replace them. I also believe AI will benefit business, rather than homes or more personal scenarios, in real commercial terms.
We should also design a user or customer interface that can display a lot of results, as we use AI to help us learn faster and make better decisions, but not expect a single answer. Furthermore, the requirement of data, both quantitative and qualitative, is indefinite. We need to encourage users and customers to provide data and feedback on an ongoing basis. We also need to design mechanisms to continuously capture, update and learn from new data as a living organism.
Lastly, a well-defined area of focus should be set, usually a narrow one, for the problem you are trying to solve. Do not dream of inventing a superpower technology on day one.
Have you read?
I foresee China as a critical hub in the global development of AI. It offers a number of conditions that are suitable for creating world-class companies in the AI field:
- Talent pool. Chinese researchers are already savvy in AI. In 2015, 43% of the top academic papers relating to AI were published with one or more Chinese researchers, regardless of where in the world the work was primarily conducted. Chinese people are proud of their mathematics, engineering and science training. An influx of high calibre young talent has and will become the quintessential foundation of any new industry.
- Traditional industries. Today, we meet many traditional Chinese companies lagging behind US enterprises in terms of the level of technological adoption. But these companies have data and money, and are eager to invest once AI experts present them with the opportunities to grow their business or make bigger savings.
- Internet market. China has one of the world’s largest internet markets with around 800 million connected users and many internet companies. Often when non-AI technology companies grow to a certain size, they need to explore AI in order to upgrade and scale.
- Closed yet open. Despite the fact that US companies are now leading AI development around the globe, there are visible barriers to entering the Chinese market. The Chinese market will need local solutions and providers. In contrast, policy around AI in China is relatively more open for experimentation and solutions.
With the advent of AI, humans can finally be liberated from mechanical tasks and devote their time to more intellectual, creative and productive explorations. AI technology presents the greatest opportunity in human history and I encourage you to pay attention to what is happening in China.