When you’re arguing with your significant other, best friend or coworker, it can be hard to feel like you’re being completely understood. Human communication is a messy and imperfect art, marred by lack of social grace, limited amounts of empathy, and consistent aversion to admitting wrongdoing. This messiness is all part of being a human, and navigating human relationships. The question is, can Artificial Intelligence handle it any better?

AI researchers and specialists seem to think so. Consider the social cues we pick up on when communicating with one another that go beyond the literal meaning of the words we speak. Placing stress or emphasis on a certain part of speech can dramatically alter the inferred meaning of a sentence. (How are you doing? Becomes How are you doing?)

 Researchers have studied how humans pick up on and ignore certain, more complex, social indicators and how that skill affects the smoothness of our communication. Humans can understand the different implications of stressing certain parts of speech, but when it comes to detecting more acute changes and indicators in human communication, AI might be better equipped to handle it.

 One study pitted a machine learning algorithm against a group of trained therapists and relationship experts to predict the outcome of over 100 couples who sought counseling over two years. Not only was the algorithm more likely to be correct about which couples would stay together (79.3% accuracy), the algorithm was also able to pick up and collect data on metrics of human speech, focused on rhythm, volume and vocal quality, that are too advanced for our regular perception.

 How do we contend with and make use of this new reality, where computers can understand more about human relations than actual humans do?

For starters, we can use AI technology to our advantage when building and maintaining relationships. Researchers have used machine learning research to validate and underscore the idea that humans are constantly projecting more information about our thoughts and feelings than we can even be consciously aware of. After letting the algorithm collect data on the couples, a model through which an AI relationship therapist could operate started to become clear.

 Researchers at Texas A&M have been developing conflict detection AI software, that can pick up on the subtle indicators that trouble is brewing that might lead to an argument. The non-invasive sensors worn by subjects detect and measure vocal indicators that the previously discussed therapist-competing algorithm had a discrete awareness of. They also pick up on physical indicators perspiration and heart rate, all of which can indicate a possible impending argument.

 What results is a preventative approach to conflict resolution that is less fraught than the role of a marriage and family counselor or a relationship therapist, or even simply a friend acting as a mediator. The role of the AI based conflict detection software is not to get in the middle of your conflict, but stop it before it even begins. The software doesn’t need to know a ton about the intricacies of your relationship. Instead, it detects and analyzes various warning signs that we often overlook when escalating a conflict. What results is not exactly conflict resolution, but conflict avoidance. Think of it like a notification on your phone that tells you to think twice before heading into a heated argument. Your relationship might truly be better for it.

 Human beings will never be perfect communicators, but many of us would like to learn how to become better communicators with the people in our lives. When faced with a conflict, it might better not to give into anger and instead modulate your emotions in a way that is healthier for all parties involved. AI conflict detection software and machine learning algorithms that are designed to detect and interpret the subtleties of human interaction and emotions, can be used by humans to become more sensitive communicators in the future. And wouldn’t that ultimately be better for all our relationships in the long run.