From schools to medical care, all sectors amplified technology usage due to the pandemic. As with anything, there are good, and evil’ points to artificial intelligence (AI). Certainly, so when it comes to emotional AI, which is the focus of this article. CXM aims, through this comprehensive guide, to help you better understand what emotional AI is. Such as how it can impact CX as well as the drawbacks of this growing and developing technology, and a lot more. So, let’s dive in.
What is emotional AI?
When it comes to learning something new, you must start at the beginning. AI is when technology learns and adapts to inputs that it receives. It is a technology developed with the idea of obtaining problem-solving and decision-making capabilities such as that of the human mind. Emotional AI is a sub-set of this.
It tries to understand and interpret human expressions, and this means verbal but mostly visually. Emotional AI mainly focuses on facial recognition, with the human face being the only observable object. That is why emotional AI uses many cameras that detect and analyse facial expressions. It is the science behind technology processing, interpreting, recognizing, and even simulating emotions.
Furthermore, it is also referred to as Affective Computing. The latter term was coined in Rosalind Picard’s paper, “Affective Computing: Challenges”. The paper was published by MIT Media Laboratory in 1995.
Since this paper was published, researchers and innovators jumped at the chance to develop methods that are connected to DeepTech solutions. This was done as a way to quantify human feelings and moods. Furthermore, biometric sensors, text, voice analysis, as well as computer vision, have also been included in data collection to create emotional AI.
According to Kate Crawford, the book author of Atlas of AI, companies have captured a lot of surface-level imagery of human expressions—billions of Instagram selfies, Pinterest portraits, TikTok videos, and Flickr photos. Therefore, we all should be aware of our participation in emotional AI development, consciously or subconsciously.
How does emotional AI impact CX?
CX (customer experience), as you know, is all about getting to know people and their wants. However, with people come emotions. Therefore, it is not much of a surprise that many businesses have, or are starting to, implement emotional AI to transform their CX to try to serve people better.
But the main way that emotional AI can help the CX industry is if the business side analyses the data from the emotional AI in real-time. It is here where the companies can see how people genuinely feel about their products and/or services, whether happy emotions or angry and bitter ones.
The AI can figure out the latter so that companies can develop strategies for improvement. While knowing the former, lets businesses know what to keep and maintain the aspects that work. Even attempting to make these points better. Furthermore, emotional AI can also speed up the work process by letting minor issues be handled by the AI, while more intense ones (that the emotional AI is not familiar with) can go to the employees to handle.
Emotional AI achieves this by using cameras, as we mentioned, to perform facial coding as well as eye-tracking. These are the features that display the most emotions. Therefore, businesses can collect the data to measure customer engagement with the hope of improving their CX and customer service. Here are detailed examples of how emotional AI is being used by big brands to improve their CX.
Emotional AI examples
As we mentioned at the beginning of this guide, emotional AI is used both for good and ‘evil’ purposes. Think of Facebook analysing our facial expressions, tracing how much time we spend searching for a product and advising companies to increase prices according to how bad we want or need it. Makes you wonder, doesn’t it?
However, emotional AI can also be applied to help cure depression, prevent suicide, or facilitate the development of children with autism. For instance, in 2017, the Nature Human Behaviour journal published research showing that the AI system was able to identify those known to be ‘suicidal’ with an accuracy of 91%. But a small sample (34 young adults) who participated in the research can’t represent and talk in name of people across different geographies.
This further indicates that medical scientists and technologists need to join their forces and get funds to explore the potentials of AI in maybe fewer commercial applications essential for a prosperous society.
Emotional AI and personalization in CX
Let’s now look at the example of Spotify as a technology that revolutionized music consumption with the help of people voluntarily sharing their data and preferences. How is this related to emotions, you might wonder? Well, you’ll agree that music is never only a sound, but a memory, an emotional journey through time. By applying emotional AI Spotify algorithm decides where that journey will take us.
Can humans surprise Spotify with their unpredictability?
With over 160 million users, Spotify is probably one of the most famous and loved AI-driven companies. They manage to gather a great deal of data on our feelings; music tastes, listening habits and use it to create playlists people will find authentic.
For instance, after a listener spends just enough time exploring music on Spotify, their algorithm could recommend personalized playlists based on the previous searches and time spent listening to songs or albums. They do this by conducting a serious semantic analysis of all keywords in the songs, blog posts, music reviews, people’s comments, etc. Netflix and Amazon also apply similar personalization and prediction platforms.
From the sociology point of view, there are wider implications of this technology. Before Spotify and Netflix, we had to rely on our curiosity, communication, and ‘forum exchange’ with other music and movie fans. Finding good music and movies was an adventure that could end up with disappointment but also satisfaction and reward.
With this argument, we’re again circling back to the point that AI-powered technologies can be misused by forcing people into certain identities without leaving them the space to improvise, as well as surprise themselves through accidental and spontaneous discoveries.
So, to answer the question: “Can humans surprise Spotify with their unpredictability?” the answer is yes at the beginning, but the more you use Spotify, the more it “reads you” and gets to know you. Therefore, turning the answer into a ‘no’.
Top 3 benefits of using emotional AI in CX
While emotional AI is a new world for CX, it has many benefits. We have found three of them, but there are a lot more. Emotional AI can help the CX industry by:
- Providing more accurate and authentic data due to facial coding as well as eye-tracking. It is more real because it is 90% more accurate when reading and mapping non-verbal elements of human behaviour that people are unaware of. Furthermore, this type of data collection happens in real-time, making the data current.
- Real-time data collecting can help businesses dive into actions straight away, allowing for agile decision-making.
- Another highlight of using emotional AI is that some customers prefer to talk to virtual avatars. This is what the USC study found out. Although, mainly for medical purposes, this thinking can be applied to CX. For example, it can be handy for customers who have what they view as simple issues that they cannot sort out themselves and are embarrassed about it. Such as buying a product online, for example.
Regarding the last point, emotional AI does not judge. People, regardless of how highly trained they are, often may not be able to help themselves. While many people do want to speak to human representatives, this is mainly to do with complex issues that are not on the surface. Interacting with an emotional AI can make more customers come forward to sort out the more minor concerns. Businesses still have to take note of these minor issues.
Rounding Off
Emotional AI can be a great benefit to improving the CX industry. But it is not the fix-all solution. A lot is still being discovered and developed with this form of technology. Think of it this way; emotions are quite complex. Therefore, emotional AI needs time to understand, adapt and respond. It would seem that the more the CX industry uses emotional AI, the more the AI will adapt. As long as the people “training” the system has a strong sense of morals and accounts for diversity.