If an intelligent machine has the capacity to feel, then can Artificial Intelligence have emotions? Unlike humans, artificial intelligence cannot generate emotions from introspection. For instance, it would not be able to feel love or jealousy. It would also not experience any other emotional state. Until it develops a strong emotional sensitivity, it will be unable to experience feelings. A person with autism will not experience any type of emotions, whether they be negative or positive.
While AI is increasingly being trained to interpret human emotions, it is still not sophisticated enough to understand cultural differences in how people feel. Hence, it can’t differentiate between the different meanings of smiles. Often, it makes the wrong decision. For example, it may misinterpret a smile as a sign of politeness when the true meaning is to ask for help. Its unintended emotional bias can perpetuate stereotyping.
In response to Tomkins’s research, Ekman received a large grant from the Department of Defense (DARPA) to continue his work. His studies proved that human affects are universal, which is a key factor for building emotion-recognition systems for machines. Tomkins’ experiments were also copied. To test the subject’s emotions, Ekman used photographs of human faces.
In addition to these human-like features, AI designers are also using machine learning methods to teach machines how to recognize emotions. For example, deep learning models are capable of interpreting speech using pictures or speech recognition systems. In addition to this, researchers are also experimenting with skin temperature and heart rate to detect the presence of emotions. The use of these two sensors is practical for developing smart wearables. Further, deep learning algorithms are able to distinguish between literal and non-literal statements.
As processing power increases and data is available, artificial intelligence may be able to emulate human emotion in the near future. Until then, AI will be limited to recognizing a limited range of emotions. That is, it will not experience all emotions. The research is ongoing, but it is certainly promising. The potential impact of artificial emotional intelligence on different industries is immense. The benefits of this technology are endless. It can help humans and machines in different sectors.
Businesses have traditionally used focus groups and surveys to gather insights from consumers. But emotional AI technology has the ability to capture real-time emotional reactions and decode facial expressions. Because humans are difficult to read, they often express feelings differently from how they really feel. This could make the AI prone to bias and resulting in biased results. Therefore, companies need to be careful in their development of emotional AI. If they fail to do so, the results will undermine the promise of emotional AI.
While we can’t yet guarantee that artificial intelligence will have emotions, it is becoming increasingly important to research these issues. Affectiva, a company based out of the MIT Media Lab, has been developing several AI applications to measure and understand emotions. These machines could be used in detecting distracted drivers on the road and measure the emotional reactions of consumers to advertising. The company has created the world’s largest database of emotional reactions. This database contains more than 10 million human faces from 87 countries. It was also hand-labeled by crowdworkers in Cairo.
In the 1960s, Joseph Weizenbaum created ELIZA, a natural-language processing program that repeated the words of its users in a simplified psychotherapy simulation. In 2007, SimSensei, a chatbot that responds to emotions, was developed. The program was originally created for military personnel to counsel them. Today, there is a free chatbot called Woebot that provides cognitive behavioral therapy to people via chat.
Emotion AI uses machine learning algorithms to understand human emotions. It can recognize the inflections of a person’s voice, which elicits anger or stress. It can also analyze images and pick up subtle facial expressions. Emotion AI has many uses in the industry. One company that uses the technology is Affectiva, which specializes in advertising and automotive AI. Their clients include 25 percent of the Fortune 500.