• Dr. Timothy Smith

Does Artificial Intelligence Pick Better Mates Than We Do?


Matchmaker, Matchmaker, Make me a match, Find me a find, Catch me a catch, Night after night in the dark I'm alone So find me match, Of my own. -Fiddler on the Roof

The beauty, the agony, the struggle, the chase, the process of finding a mate defines a central aspect of the human experience. Naturally, this central focus of our lives has also been the subject many of our most favorite stories, novels, books, movies and plays. These stories with their infinite variations of plot often begin with how people meet--the arranged marriage, the chance encounter, the blind date, the friend’s recommendation, the meeting of opponents with a certain chemistry, the match maker. Bringing two people together often takes a mix of luck and perseverance. Think of Cinderella meeting the prince by enchanted means or the mixture of media formats such as radio and snail mail in Sleepless in Seattle or email in You’ve Got Mail. The classic musical Fiddler on the Roof employs a matchmaker named Yente playing the central role in setting off the argument as to whether a girl should be able to choose her own husband. Regardless of how people meet, there needs to be a spark and to make this successful. Human matchmakers consider many factors when making a match from temperament, ambitions, beauty, attitude, religion, socioeconomic status and such to make the best match. According to Carrie Ritchie of the Indianapolis Star, human match makers remain popular with singles in the United States. However, the rising star in match making is the artificial intelligence that lies behind the scenes in internet dating sites.

Online dating benefits from the sheer number people that you can check out online, which far exceeds what can be done in person at a gathering, party, out at a club or at a coffee shop. Apart from the possibility to virtually meet hundreds of pre-selected potential partners online from the comfort of your own home, another aspect of online dating that the user cannot see is the use of powerful artificial intelligence and machine learning to suggest connections—in essence a computer matchmaker—a mechanized Yente. Most sites claim to have proprietary computer programs that match users based on answers they gave in the detailed questionnaires filled out when signing up.

The computer starts the process by crunching a client’s preferences and personality profile and finding a match based on similar interests and traits. Additionally, as users select people to contact the machines learn more about one’s preferences. Using your personal information and preferences, the suggestions of who you might date become more refined. The method is called machine learning. In some respects, it is the same as when you see Amazon or Netflix recommend products and movies based on what you and other people who bought what you are buying. Your purchasing patterns will match you with people like you. Online dating works in a similar way.

Online dating ranks as the second most popular way to meet people in the US behind the good old fashioned way through friends and family. The variety of dating sites continues to expand from the well-established eHarmony and Match.com to Tinder and Bumble. Over 40 million Americans have used an online dating site and the trend is going up across all sectors and age groups with significant increases in both the younger ages from 18-24 (22%) and in the older set from 55-64 (5%) according to the Pew Research Center. Clearly, internet dating continues to rise in popularity, but the question remains as to the success of internet dating. If you consider successful to mean ultimately a marriage, then internet dating has continued to improve.

A study conducted at the University of Chicago and published in a paper titled, “Marital Satisfaction and Breakups Differ Across Online and Offline Meeting Venues,” concluded that more than one third of all marriages in the US started online between the years of 2005 and 2012. The study went on to say that the marriages that began online were in fact less likely to end in divorce. Online marriages over the length of the study stayed together 94% of the time versus 92.4% of the time for analog matches (matches made the old-fashioned way). Additionally, online marriages reported a higher satisfaction score than marriages from other venues. Now it must be noted that eHarmony funded this study, but the authors claimed to have had full independence in data collection and analysis. An article in UChicagoNews.com by William Harms speculated as to the cause of difference in marriage success rates in the University of Chicago study such as noting that the population online may be more focused on finding a marriage partner than the general public in offline venues. Although not explicitly connected to the growing success of online matchmaking artificial intelligence may be a leading factor.

Already the online dating companies at their hearts are technology companies that use massive computing power to support the millions of singles looking for love across the nation. I cannot help but wonder if the modest improvement in matchmaking seen in online dating over other venues will continue to grow. Machine learning performs better with more data. Perhaps people will be willing to give up more of their data than just personality questionnaires; maybe also sharing your spending habits or data from your FitBit will make even better matches. Given the current trend of increasing data, we should anticipate computers to be better and better matchmakers with an appetite for more and more data.

Dr. Smith’s career in scientific and information research spans the areas of bioinformatics, artificial intelligence, toxicology, and chemistry. He has published a number of peer-reviewed scientific papers. He has worked over the past seventeen years developing advanced analytics, machine learning, and knowledge management tools to enable research and support high-level decision making. Tim completed his Ph.D. in Toxicology at Cornell University and a Bachelor of Science in chemistry from the University of Washington.

You can buy his book on Amazon in paperback and in kindle format here.