On July 9th, Rakuten Institute of Technology participated in Data Analytics Forum 2018, a forum focused on how can we make more effective data analytics. Among the many speakers, Masaya Mori, the global head of Rakuten Institute of Technology, held the stage for a session on Rakuten’s new efforts to better utilize data for use in Creative AI. The session also covered topics such as the status of AI, trends in deep learning, issues relating to the era of individualization, and how the internet changes the markets.
Below you can read some key takeaways from the presentation.
A big problem in the area of data analytics is lack of data, but more recent solutions are finding ways around this. Even when there is little data available, Generative Adversarial Networks can create data in an artificial way, allowing for deep learning to be used in cases where it otherwise wouldn’t be possible.
Another interesting recent development is in Creative AI, where systems are developed to do creative work like writing stories and making music. This is a major shift, as it means moving from AI that learns to AI that is able to create on its own.
Many of these AI solutions are still relatively early in their development, but before they reach higher levels of performance, if we don’t actively use AI we will not be able to survive as a company. That is especially true for the retail and consumer industries, and that is why we are putting a lot of effort into AI.
A theme we are putting a lot of emphasis on is that experts will be defeated, as it is becoming possible for even unexperienced people to produce results that are not inferior to experts. Related to this is also the phenomenon that, by looking at purchasing behavior, you can see that consumers are also getting more and more individualized and we need to respond to this.
Up until now, when considering product planning, we were thinking that if we were the customer, we would like to buy these products. But nowadays, there might be hundreds of people around Japan that would buy or already buying a product, but they are just not showing up in your vicinity. However, on the internet, distance is irrelevant. The reach of consumers is global, and before you know it, your products might be selling to people all around the world. To respond to this kind of acceleration of diversification, AI is an indispensable technology.
AI can process data stored in a fixed framework at a tremendous speed that human beings can’t handle, and get results with great precision. By contrast, humans have the power to change the framework. In other words, our business is to change the customers and to change the framework. We believe that, to combine these characteristics of humans and AI, we will need to develop new drivers for our businesses.
For the full report in Japanese, see the link below.