While artificial intelligence (AI) technology is already helping to create the smart cities of tomorrow, emotional AI is in its infancy. Able to analyze people’s emotional states, or by incorporating elements of human emotions into the technology itself, so that it can better adapt and respond, will emotional AI always make the right decisions?
It’s early days for emotional artificial intelligence (AI), but the technology is already finding its way onto devices in our home. Where Alexa, Amazon’s voice assistant, used to just play music on demand or relay the weather forecast, it now responds to people in an ‘excited’ or ‘disappointed’ tone, each of which comes with three levels of intensity, determined by the emotions it picks up from the person it is conversing with.
It’s not just chatbots – more and more areas of society are trialing and deploying emotional AI technology. Here are some of the most exciting.
Conducting market research
US company Affectiva has been using a combination of facial recognition and AI technology to support market researchers in testing people’s emotional responses to adverts since 2009. The technology monitors the physical reaction of hundreds of participants to provide marketeers with the data they need to adapt content accordingly.
Improving physical safety
Researchers and companies are developing technologies that can detect signs of fatigue in an individual and subsequently alert them to avoid a harmful incident. For example, AI technology could monitor the faces of employees in a warehouse and, upon detecting signs of fatigue, could suggest the individual takes a break before otherwise making a mistake.
It’s already being happening in the automotive market, where car manufacturers are integrating the technology into their connected cars. Camera sensors can monitor the driver’s eye openness, blink rate, head position and head angle if it detect signs of drowsiness . If the emotional AI detects signs of drowsiness, it can then alert the driver by playing loud music or changing the temperature in the car.
Supporting mental health
CompanionMx is a digital health technology company with a smartphone app that uses emotional AI for proactive mental health monitoring with the aim of achieving better clinical outcomes. In 2019, the app was selected by the US Department of Defense for a suicide prevention study in active duty navy personnel.
Deployed in a smartphone app, the emotional AI actively monitors the user’s voice and monitors background smartphone data, as well as asking the individual questions about their wellbeing via the app, to produce markers of emotional wellbeing. These predict some of the core symptoms of mood and anxiety disorders: depressed mood, diminished interest, avoidance and fatigue. The data can then be shared with clinicians, while the user will receive alerts that prompt them to take action, such as breathing exercises.
Staying on the call
US-based tech company Cogito uses AI to provide in-call voice analysis and real-time guidance to enhance behavior for those working in call centres and telesales. The software analyzes the call’s energy, interruption, empathy, participation, tone and pace, and guides the employee to speak withmore empathy, confidence, professionalism and efficiency where needed.
Ethics of emotional AI
Emotional AI technology currently relies on data that comes from a tiny slice of the population, whereas in reality emotion is expressed differently by individuals across the globe, particularly between cultures. As the technology evolves, it’s crucial that there is no bias behind the technology. Developers must therefore draw from diverse data sets – which remain secure and protected – and which encompass different countries and cultures, when creating the algorithms that will determine what certain types of behavior signify.
Providing the technology can be ethically and transparently deployed, the applications are limitless – from supporting people with mental illness to keeping individuals safer in the workplace and on the roads.