Casual Dating With The Help Of Artificial Intelligence| ItSoftNews

HowStuffWorks Now: Tired of Swiping Left? HowStuffWorks HowStuffWorks Now: Tired of Swiping Left?

Since it was first tested on the campus of UCLA, Tinder has become incredibly popular. As we briefly explain in the video, the app pulls profiles of people in your general area and lets you flip through them quickly, swiping right if you think the person is attractive or left if you’re not interested. It’s the digital equivalent of approaching someone in a real-world setting because you find them attractive.

Tinder lets you send messages to the people you’re interested in only if they too have swiped right on your profile. That way, users can avoid receiving unwanted messages from random people. But let’s say you have particular preferences in physical appearances. For example, you find bald guys with a goatee to be irresistible. You might get tired of swiping left as you pass by guys with full, lustrous manes of hair, sighing unhappily all the while.

Advertisement

That’s where artificial intelligence might come in handy. Computer scientist Harm de Vries decided to streamline the Tinder experience by creating an AI program capable of learning de Vries’ preferences. The software could dismiss profiles that clearly didn’t fit what de Vries thought of as attractive, leaving behind a pool of potential dates.

This isn’t as easy as it sounds. Harm de Vries had to feed thousands of profiles into the program to teach it how to identify the traits he found attractive or unattractive. Learning algorithms like this one have become popular in the artificial intelligence field, where we’re seeing scientists teach computers rather than program them. IBM’s famous Watson system uses a similar approach to aid doctors in diagnosing patients. It’s serious stuff.

So while this particular story might seem insubstantial, underneath is an important concept. We’re entering a world in which machines will be able to learn, which means they’ll also be able to improve over time. A programmed machine will always perform a task the same way. A learning machine will find ways to perform tasks more effectively. And if that helps you score a few dates, that’s awesome too!

Advertisement

Leave a Reply

Your email address will not be published. Required fields are marked *