Blog Posts — May 7, 2018

These Robots Will Recognize Your Moods

Advancements in artificial intelligence (AI) have continued apace over the last few years. Thanks to improved connectivity, faster processing speeds and astounding developments in machine learning, AI is now part of our everyday life.

A recent PULSE ONLINE article reported on Accenture’s forecast¹ that AI will double the pace of growth in 12 developed countries by 2035. What’s more, according to a recent PwC survey, executives think that AI will be crucial for their success with 72% believing it will be the business advantage of the future.²

With the launch last January of the STOXX® AI Global Artificial Intelligence Index, the first thematic index using AI technology, STOXX introduced a channel to access a broad and geographically diversified range of AI-related stocks. Companies eligible for the index are those that invest heavily in the development and adoption of AI-related technologies.

This year worldwide spending on cognitive and AI systems may increase 54% to $19.1 billion, and grow to $52.2 billion in 2021, according to the International Data Corporation (IDC).

At the cutting edge of AI development, cognitive technologies are a prime example of tasks that only humans used to be able to do. We are already familiar with cognitive concepts such as computer vision, natural language processing, speech recognition and robotics. However, a new category of cognitive technologies is emerging, that of emotion AI.

Making productive use of emotion AI

Artificial emotional intelligence, or emotion AI, is a set of programs that can detect more complex human feelings than happiness or anger – something already achievable through facial recognition. Using computer vision and deep-learning techniques, these programs identify and classify a much wider range of facial and vocal expressions. One use for this type of programs would be tracking the emotions of people queuing at airports in order to find out if this affects their mood for spending.

IBM Corp., the global cloud computing and analytics company, is working on training computers to make sense of emotions such as distraction, confusion, fatigue, attention and even nostalgia, and to directly translate them to AI.³ Examples include the use of an AI confidante that deals with workplace issues or digital assistants that could listen in on customer service agents’ phone calls to measure and predict client moods.

Japan’s Softbank Robotics has designed a robot called Pepper, which can interpret human emotions and act as a day-to-day humanoid companion. Pepper takes a person’s voice, the expression on their face, body movements and words spoken to interpret their emotion and offer appropriate content.

Accessing an investable AI product

Equity investors looking to participate in these pioneers’ growth have faced huge challenges. The breadth and scale of AI companies that exist, both AI adopters and innovators, is vast, and for many investors it has been potentially unnavigable. While leviathan companies such as Amazon and Alphabet have provided plenty of opportunities for returns, staying ahead of the AI technology curve is getting tougher.

The STOXX AI Global Artificial Intelligence Index was designed to access exactly those companies at the frontline of AI investing. Both IBM and Softbank Robotics’ parent companies are members of the index.

Identifying the innovators

The primary criterion for selection in the index is the relative number of patent filings that an algorithm finds to be related to AI intellectual property. Research shows that AI patent applications, in particular, are on the increase. IBM filed over 700 patents last year,⁴ and Qualcomm, Microsoft and Sony are regularly among the top filers. Last year, Disney filed a patent application for a theme park attraction that senses and manages vehicle behavior based on occupant awareness.

But the innovation landscape is more diverse than would probably be assumed, according to ClearView IP, an intellectual property and consulting firm.5 While the number of patents filed overall per year by the 50 most frequent sponsors has remained relatively constant, there is a much wider presence of assignees now applying for AI patents. In 2016, there were over 1,600 different entities filing AI-related patents, four times the amount of 20 years earlier.

Companies in the STOXX AI Global Artificial Intelligence Index include hardware and data providers as well as AI users that need to file their own patents. A high percentage come from Japan, Korea and Taiwan, although currently just over half are US-based.

The STOXX AI Global Artificial Intelligence Index is an innovative and powerful approach to passive investing. Now, investors can take advantage of opportunities in the field of emotion AI and the scale and scope of technological changes that lie ahead.

Featured indices

STOXX® AI Global Artificial Intelligence Index

1 Accenture, ‘Why artificial intelligence is the future of growth,’ Sep. 28, 2016.
2 PwC, ‘Bot Me; A revolutionary partnership – how AI is pushing man and machine closer together,’ 2017.
3 TEDTalk, Raphael Arar, IBM, ‘How we can teach computers to make sense of our emotions,’ Dec 2017.
4,5 ClearView IP, Examining Intellectual Property in Growing AI Market, Feb. 14, 2017.