Main content starts here.
Japanese English

Creating value for society through Marketing × Innovation

October 28, 2016

How will the role of people change with the advance of AI?
What kinds of value will be produced when things and people are connected through IoT, providing us with a deeper understanding of contexts?

We spoke with NEC's Takaaki Shimizu (CMO), Katsumi Emura (CTO), and Kumi Fujisawa from the think tank SophiaBank, who is involved in international consulting and critique.

What must people do to produce value from AI?

“Digital” is currently a major theme when speaking of business and technology, from the viewpoint of the “integration of marketing and innovation.”

To produce digital value requires understanding of both “digitization,” converting to and using data in digital form, and also “digitalization,” which is the transformation of business through technology, and engaging in both these activities in the medium-term. It has become possible to solve business issues by creating value from data through the advanced capabilities of computing, networks, and software. Artificial intelligence (AI) is again receiving attention as one form of software. We must note that in earlier AI “booms,” AI was misunderstood as something that would take over all human tasks.

Takaaki Shimizu, Executive Vice President and CMO (Chief Marketing Officer), NEC

Unless someone decides on something concrete that they want to do, AI won't think of it for them. This is a key point. Humans recursively rethink things in an analog way, which is the antithesis of “digital,” and you could say that this is like an Eastern way of looking at things. NEC is introducing a system to monitor for potential faults in plants by collecting data from several thousand sensors installed in the plant and analyzing the data by using AI. The technology takes an overall view of things and detects potential faults by looking for breakdowns in the interrelations among sensor data. This is similar to how Eastern medicine makes a diagnosis by looking at the whole body. In that sense AI and the Internet of Things (IoT) solve problems by looking globally at a lot of data, which could be said to be more like an Eastern medicine approach. In my AI research, I've come to think people's job is to think more deeply about things. That may be because I'm getting older, though.

AI is expanding into more and more fields, such as factory automation, and self-driving cars, isn't it?

AI is able to handle an increasing number of things that people do normally and regularly. The important issue is not that cars can drive automatically, but to consider what new things people can do when they are freed from driving automobiles. From our interaction with customers, we have learned much, become aware of changes in society, and recognize that people are what is important.

Kumi Fujisawa, Moderator, Think Tank SophiaBank

“Five-senses AI” cooperating with humans

I understand that you have established the NEC AI brand.

Yes, we call it “NEC the WISE.” “Wise” is an adjective, but “the wise” is a plural noun, meaning “wise people.” So we use “NEC the WISE” to refer to NEC's various cutting-edge AI technologies, with the idea that people and AI work together on solutions to complex and advanced problems. Realizing “Visualization (of the real world),” “Analysis,” and “Prescription,” and optimizing those processes are features of NEC the WISE.

The image of a brain is strong with AI, but AI at NEC seems broader, including looking, listening, and determining actions based on the results of analysis. Like using the five human senses...

Yes, we have technologies such as speech recognition, language understanding, and estimation and prediction. For a long time, we have had particularly strong technologies for extracting data from images and interpreting it. About 50 years ago, we commercialized a machine for reading postal codes, and we have been working on biometrics technologies in the form of Fingerprint Identification and Face Recognition for more than 40 years. These are being used in more than 70 countries around the world.

Main content ends here.
Top of this page