Imagine if an app could tell you, with certainty, whether someone you’re with is angry at you, even if they’re trying to hide it. Or sad, fatigued, or in any other emotional state. So-called “Empathy Artificial Intelligence” is the latest thing in company customer service, and a lot of other applications as well.
The latest generation of AIs have come about thanks to an increase in data available for computers to learn from, as well as their improved processing power. These machines are increasingly competitive in tasks that have always been perceived as human.
Recently, the U.S. insurance company MetLife started trying out a so-called “Empathy A.I.” or digital assistant called “Cogito,” which listens in on phone calls between clients and customer service representatives. Using an algorithm, it will signal to the rep when it detects the client is experiencing emotion, and it will offer advice and tips. Similarly, it will look for signs of fatigue or agitation in the customer service agent, and make suggestions to help. Sort of like an empathy coach.
Also, researchers have developed and approach called “deep learning” where A.I. is able to tell whether a person is a criminal just by looking at their facial features. The system was evaluated using a database of Chinese ID photos and the results are truly jaw dropping. Turns out the AI mistakenly categorized innocents as criminals in only around 6% of the cases, while it was able to successfully identify around 83% of the criminals. This leads to a staggering overall accuracy of almost 90%. Given the success of deep learning AI, it is not surprising that artificial neural networks can distinguish criminals from non-criminals – if there really are facial features that can discriminate between them.
The bottom line from these two examples is that AI seems to be able to find hidden emotions from your face and subtle voice prints so minor that it might be undetectable by humans.
Subjective topics such as emotions and sentiment are still difficult for AI to learn, partly because the AI may not have access to enough good data to analyze them objectively. For instance, could AI ever understand sarcasm? A given sentence may be sarcastic when spoken in one context but not in another.
Yet the amount of data and processing power continues to grow. So, with a few exceptions, AI may well be able to match humans in recognizing different types of emotions in the next few decades. But whether an AI could ever experience emotions is a controversial subject. Even if they could, there may certainly be emotions they could never experience – making it difficult to ever truly understand them.
At this point in our development of A.I., computers can’t truly be intuitive until they understand how you feel and not just how you act. While the idea of an ’empathic AI’ sounds good while your on the phone with a frustrated client, as humans we are designed to empathize with other people in an intuitive way, picking up on subtle tone and language and relying on a computer to tell us that could make people believe that someone is frustrated with them, even if they’re only having a frustrating day.
It has good applications for people with autism or a disability that makes it particularly difficult to read emotions. And it could be used as a way to monitor, say, long distance truck drivers to measure how fatigued they may be.
Leave a Reply