An AI just predicted what makes a successful relationship
The key to a successful relationship is never easy to understand. Some would say it is trust, some would go with sexual compatibility, while others would argue it depends from individual to individual and their expectations. However, when scientists quizzed an AI engine, which analyzed thousands of relationships as part of a study, the answer was more systematic and solid. Here's what it found.
40% of US marriages end up with divorce
Today, just 1% of Indian marriages end up in divorce, courtesy all the social stigma/pressure on Indian women, but other countries have witnessed surging divorce rates of late. In the US, the divorce rate has gone over 40%, leaving questions around why so many people are ending their relationships — which in turn affects their (and their partners') lifestyle, career, productivity, among other things.
First-of-its-kind AI study of relationships
To find answers and predict what is most critical to relationship satisfaction, researchers from Western University took a tech-backed approach. They conducted the first-of-its-kind artificial intelligence study of romantic relationships under which the data on as many as 11,196 couples, taken from 43 distinct self-reported datasets, was fed into a random forest classification machine learning algorithm.
Then, the AI found predictors of successful relationships
Once the data was fed, the algorithm scoured through it and discovered patterns synonymous with a successful relationship. In the findings, the AI suggested that the most reliable predictor of romantic happiness is the partners' belief that their better half is fully committed. Beyond this, it found that things like appreciation, sexual satisfaction, and conflict resolution are also important to make a happy relationship.
Relationship-based factors outweighed individual characteristics
The AI's analysis also seemed to indicate that the aforementioned relationship-specific factors were far more critical to couple's happiness than characteristics specific to an individual, like their life satisfaction, depression, income, or agreeableness. "In other words, relationship satisfaction is not well-explained by your partner's own self-reported characteristics," Western University professor Samantha Joel, who led the study, said in a statement.
"Individual differences fade into the background"
"Relationships-specific variables were about two to three times as predictive as individual differences, which I think would fit many people's intuitions," Joel added. "But the surprising part is that once you have all the relationship-specific data in hand, the individual differences fade into the background."
So, who you are with your partner is extremely important
The researchers clarified that their algorithm's findings do not suggest that a person's choice of partner - including their individual characteristics - does not matter. Instead, they said, it just shows that who they are is much less important than who they are when they are with you for the overall success of the relationship.