Blog: Will emotionally intelligent machines enable better outcomes?
“Emotion pulls the levers of our lives, whether it be by the song in our heart, or the curiosity that drives our scientific inquiry.” — Rosalind Picard, Professor at MIT and Founder of the Affective Computing Research Group at MIT Media lab
The devices we interact with have become increasingly smart. For example, by assessing your location and the traffic to the airport, your smartphone may prompt you to leave early to catch your flight on time. However, while today’s devices display logical reasoning, they lack emotional awareness — a critical factor in decision-making as shown by studies in neuroscience.
Addressing this cognitive-emotional gap in current technology devices and experiences is the goal of affective computing. Affective computing is defined as computing that relates to, arises from or influences emotions. Proponents of and experts in the field, founded by Rosalind Picard at MIT, assert that emotional intelligence is fundamental to achieving meaningful human-machine interactions. In the above example, an intelligent device might know that you have a fear of flight and along with reminding you to depart for the airport, it might soothe your anxiety with encouraging words and empathy much like a friend would.
To be clear, affective computing does not seek to give machines the ability to “feel” emotion, but rather to decipher and appropriately respond to human emotion. By combining data from inputs such as, facial expression, voice inflection, skin temperature, blood flow and body language, and employing machine learning, affective computing hopes to enable more natural and intelligent interactions with technology.
In the future, devices will not just deal in facts, they will also factor in emotions.
Emotional intelligence: coming to a device near you
While this might sound like science fiction, in reality affective devices are steadily making their way into our offices, schools, movie theaters and into our homes.
In professions where decisions carry great risk, such as stock trading, the emotional well-being of employees is particularly important. Banks are experimenting with systems to monitor the emotional state of their workers through sensor-laden badges that transmit physiological data for managers to analyze and guide when necessary. Elsewhere corporate brands and marketing teams are using affective software to gauge the effectiveness of ads, movies and TV shows.
Emotional intelligence is also being used to build trust in and bolster the social acceptance of future technologies. One start-up is hoping to alleviate anxieties around autonomous vehicles by teaching self-driving cars to detect the comfort level of its passengers and adjust its driving style accordingly.
Similar examples exist in education to improve student engagement and personalize teaching. In health care, affective devices are not only enabling better patient outcomes, they are also informing research in areas such as autism, depression and stress.
With one London-based startup developing an “Emotion Processing Unit” or EPU that can be embedded in any compute-enabled device, we can expect to have many such sentic (or affective) experiences in the future.
The obstacles to achieving affective machines
Today, affective devices are in nascent stages. Much like the current state of AI, applications of affective computing are narrow. And there are significant practical, technological and ethical challenges to fully realizing the vision of emotionally intelligent computing.
Perhaps the foremost obstacle is the primitive understanding we currently have of the human brain. Although there is no broad consensus on the connection between emotions and physiological response, there is agreement that a link does exist. But, correlating physiological cues with one’s emotional state is complex and varies by individual, context, gender and culture. Specific situations may not be emotional for everyone and one might not be equally emotional in all situations. Moreover, different individuals may exhibit different physiological responses to the same emotional state.
Monitoring, collecting and synthesizing data from various modalities — voice inflection, skin temperature, body language etc — is another challenge. For instance, electrodes placed on the scalp that capture brain waves would be ideal for capturing one’s neural patterns in various emotional states. However, consumers and professionals would find such a device invasive and uncomfortable.
Machines are already privy to so much information about us — financial and health information, shopping preferences and web-browsing patterns — our emotions are seemingly the last vestige of privacy. Collecting individual physiological and emotional data will inevitably raise certain ethical questions. Who will own this data and how will its use be regulated? The availability of such information could lead to manipulation. For instance, consumers might feel maneuvered by brands employing affective computing to nudge them towards purchasing one product over the other.
The bigger picture
With the addition of emotionally intelligent devices, our environment will not only serve up data-driven recommendations and aid our decision-making, but also tailor its interactions to suit our mood and emotional state. In doing so, affective computing has significant implications for several megatrends that are transforming the human experience.
Societal challenges such as climate change have become increasingly urgent. Beyond generic nudges, compute-enabled devices may use emotion to motivate behavioral changes that yield both individual and societal benefits. For instance, you might be more likely to recycle if you are shown the impact it will have on the future world that your children and grandchildren will inhabit.
Tthe evolution and proliferation of technology augurs a future of frictionless markets and experiences. The super consumer of the future will expect seamless access to products and services as well as bespoke experiences tailored to their specific context and requirements. Affective computing can take this vision even further, allowing companies to craft personalized and emotionally-relevant consumer experiences. However, issues of surveillance and privacy will assume even more importance, thus emphasizing the role of behavioral design.
The automation of work has dominated social and political narratives across the world. Many contend that while automation will lead to displacing routinized jobs, it will not displace jobs requiring emotional intelligence. Consequently, affective computing may play a central role in helping employees develop better emotional intelligence — a key differentiating skill for humans versus machines in future labor markets.
On the other hand, if machines become adept at both logical reasoning and emotional awareness, they may eventually displace humans in the workplace entirely.
Emotion capture as the future basis of competition?
Machines that decipher and respond to human emotion have long populated science fiction. Now they are emerging from labs into the real world and promise a more meaningful paradigm for human-machine interaction. However, to realize this vision, affective computing will need to overcome a number of hurdles — including technology and design challenges, and issues of privacy and ethics.
Companies will need to balance using customers’ emotional states to improve user experiences against privacy and security concerns. Governments will need to weigh the benefits of more effective “nudges” against the ethical concerns of gathering citizens’ emotional data. And consumers will need to consider the trade-off between enjoying more personalized experiences and giving up their most private feelings to obtain them.
For the moment, it is altogether too early in the development of the field to accurately foresee the transformative impact of emotionally intelligent digital experiences. Regardless of how the field evolves and when it becomes mainstream, one thing is clear — capturing human emotion and wielding it effectively may become the future basis of competition and it will fundamentally change the human experience.
I lead technology insights and thought leadership at EYQ, EY’s global think-tank that convenes and curates diverse perspectives from across business and academia. The better question which drives our research, informs our analysis and inspires us is: What’s after what’s next? The views in this article are mine and do not necessarily reflect the views of the global EY organization or its member firms.