Date
16 October 2018
New technologies, such as programs that allow machines to learn from each other, are likely to lead to breakthroughs in medicine and many other areas. Photo: Reuters
New technologies, such as programs that allow machines to learn from each other, are likely to lead to breakthroughs in medicine and many other areas. Photo: Reuters

How machines can get smarter more quickly

The rise of internet technologies has inspired us to think about future trends. I’m fortunate to be involved in many innovative technology projects, and here are a couple of observations I’d like to share.

First is about data security risks. Cases of data leaks and online fraud are rising. That being the case, new methods of identity verification and privacy protection are likely to emerge and reshape the business world.

For example, blockchain technology offers authentic, encrypted and easily verifiable records.

Shopin uses blockchain technology to build a universal shopper profile, where merchants can store customer data. The data can only be decoded and shared with other retailers after being approved by customers.

Blockchain has tackled the trust issue and brought convenience to data sharing. This decentralized technology is likely to see major breakthroughs in the next 18 months.

The second trend is about artificial intelligence. Do you want your AI program to be smarter? Let it compete with other AI programs. Generative adversarial networks (GANs) make this possible.

GANs consist of two neural networks that are pitted against each other. They can be applied to any field or business to help machines learn quickly. They have wide applications and are likely to lead to breakthroughs in medicine, transport and many other areas.

The full article appeared in the Hong Kong Economic Journal on July 25

Translation by Julie Zhu

[Chinese version 中文版]

– Contact us at [email protected]

RT/CG

Venture Partner of Sequoia Capital China, former head of the data committee and vice president at Alibaba Group.

EJI Weekly Newsletter

Please click here to unsubscribe